Submission to the

National Transportation Safety Board for the

Investigation of Colgan Air Flight 3407 Accident Bombardier Dash 8-Q400, N200WQ Clarence Center, New York, February 12, 2009

December 4, 2009

Mike Crook Director of Safety Party Coordinator Colgan Air, Inc.

Table of Contents Page I.

Factual Information 1.1 History of Flight 1.2 Damage to Aircraft 1.3 Other Damage (Ground Damage) 1.4 Personnel Information 1.5 Aircraft Information 1.6 Meteorological Information 1.7 Aids to Navigation 1.8 Communications 1.9 Flight Recorders 1.10 History of Colgan Air, Inc. 1.11 Organizational and Management Information 1.12 Flight Schedule of Flight Crew 1.13 Captain Renslow’s Checking and Training Prior to Joining Colgan Air 1.14 FO Shaw’s Checking and Training Prior to Joining Colgan Air 1.15 Colgan Air Hiring Process 1.16 Pilot Records Improvement Act of 1996 (PRIA) 1.17 Freedom of Information Act (FOIA) 1.18 Hiring History of Captain Renslow 1.19 Hiring History of FO Shaw 1.20 Colgan Air Q400 Training Program 1.21 Captain Renslow’s Checking and Training at Colgan Air 1.22 FO Shaw’s Checking and Training at Colgan Air 1.23 Colgan Air Stall Warning Training 1.24 Flight Crew’s Stall Warning Training 1.25 Colgan Air/Q400 Winter Operations Procedures 1.26 Q400 Airspeed Indication, Speed Bug and Low Speed Cue 1.27 The Q400 Reference Speeds Switch 1.28 AeroData Calculations for Proper Landing Speed and Flight 3407 Speed 1.29 Colgan Air Sterile Cockpit Policy 1.30 Colgan Air Aviation Safety Action Program (ASAP) 1.31 Colgan Air Fatigue Policy 1.32 Colgan Air Pilot Scheduling 1.33 Colgan Air Commuting Policy 1.34 Colgan Air Sick Leave Policy 1.35 Colgan Air CRM Training 1.36 Colgan Air Safety Culture

1 4 5 6 8 8 11 11 11 11 12 13 15 17 18 18 19 19 20 20 22 25 26 28 31 34 34 36 37 38 38 40 41 41 42 43

II.

Analysis 2.1

2.2

2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10 2.11 2.12 2.13 2.14 2.15 2.16 2.17 III. IV. V.

The Flight Crew Did Not Respond to the Unexpected Stall Warning in Accordance with Colgan Air’s Training and Prescribed Procedures The Lack of Warnings Regarding the Effect of Setting a Non-Ice Vref Speed with the REF SPEEDS Switch Set to INCR Led to the Stall Warning The Lack of a Warning or Caution Light Alerting the Flight Crew to the REF SPEEDS Led to the Stall Warning The Flight Crew’s Sterile Cockpit Violations May Have Caused a Loss of Situational Awareness The Flight Crew Did Not Adhere to Colgan Air Policy on When Checklists Should Be Accomplished FO Shaw Did Not Properly Manage Her Personal Schedule Preceding the Accident Captain Renslow May Have Not Properly Managed His Personal Schedule Preceding the Accident Colgan Air Has a Careful, Robust Hiring Process Colgan Air Has a Through and Effective Pilot Checking Program Colgan Air Q400 Training Program is Robust Colgan Air Provides Appropriate Remedial Training Colgan Air Provided the Flight Crew with Proper and Thorough Stall Training Colgan Air Provides Appropriate Winter Operations Training FO Shaw Had Substantial Winter Operations Experience Colgan Air Pilot Scheduling Affords Ample Time for Pilot Rest Colgan Air Fatigue Policy Promotes Safety Colgan Air Commuting Policy is Not a Factor in the Accident

Probable Cause Contributing Causes Recommendations

46

46

47 48 50 50 51 52 53 53 55 56 57 59 62 63 64 65 65 66

Executive Summary On February 12, 2009, at approximately 22171, a Colgan Air Bombardier Dash 8-Q400, N200WQ, operating as Continental Connection Flight 3407, crashed approximately five nautical miles northeast of the Buffalo-Niagara International Airport during an instrument approach to runway 23. The four crewmembers and 45 passengers were fatally injured and the aircraft was destroyed by impact forces and post crash fire. There was one ground fatality. The flight, from Newark, New Jersey, to Buffalo, New York, was operating under Title 14 Code of Federal Regulations Part 121. The probable cause of the accident was the flight crew’s loss of situational awareness and failure to follow Colgan Air training and procedures, which led to a loss of control of the aircraft. Contributing to the accident were: 1. The flight crew’s failure to follow Colgan Air procedures and training regarding the proper response to a stick shaker. 2. The lack of an adequate cockpit warning system in the Q400 to warn the flight crew when a Speed Bug is set to a speed below the calculated stall warning speed. 3. The lack of an adequate warning in the Q400 Aircraft Flight Manual or Aircraft Operating Manual regarding the effect of setting a non-ice reference speed with the REF SPEEDS switch set to INCR during approach and landing. 4. The flight crew’s non-pertinent conversation during the descent and approach phase, in violation of Colgan Air’s training and procedures.

This submission discusses the flight crew’s decisions and actions, the flight crew’s stall avoidance and recovery training, the Q400’s ice protection and stall warning systems, and Colgan Air’s policies and procedures regarding winter operations, pilot hiring, sterile cockpit, fatigue, and commuting. This submission also addresses Colgan Air’s safety programs, pilot checking, and pilot training. I.

Factual Information 1.1.

History of Flight

Captain Marvin Renslow and First Officer Rebecca Shaw were scheduled to report for duty at the company’s base at EWR on February 12, 2009, at 1330. See Operations Group Chairman Factual Report, p. 2. The first two flights of the day, planned as a round 1

Unless otherwise specified, all times set forth herein are Eastern Standard Time.

1

trip returning to EWR, were cancelled due to high winds that caused numerous other cancellations at EWR that day. Id. The Sabre Crew Trac Pairing Print Report for Flight 3407 showed a scheduled departure time for Flight 3407 of 1945 and a scheduled arrival time of 2221. See Operations Group Chairman Factual Report, p. 3. The company dispatch release for Flight 3407, which was issued at 1800, showed an estimated time of departure of 1910 and an estimated time enroute of 53 minutes. Id. A Departure Clearance Request was made at 1930 and an OUT (pushback from the gate) report was made at 1945. Id. The aircraft for Flight 3407 was a Colgan Air Bombardier Dash 8-Q400, N200WQ. Id. EWR Ground Control gave the flight taxi instructions at 2030, and EWR Tower cleared the flight for takeoff at 2118. See Operations Group Chairman Factual Report, p. 3. The ACARS report showed the flight OFF (airborne) at 2119. Id. The flight’s dispatch release showed an intended route of flight from EWR to COATE intersection, then via airway V-126 to LHY, direct ULW, direct BENEE, then airway V-164 to BUF. Id. The intended cruise altitude was 16,000’ MSL. Id. According to FAA records, the flight was cleared by New York Center to maintain 16,000’ MSL at 2131. Id. During the climb to 16,000 feet, all of the de-ice systems (propeller deice, pitot deice, and airframe deice equipment) were selected on and remained on throughout the flight. Id. The REF SPEEDS switch was set to INCR. See NTSB Public Hearing Transcript, p. 14. The last reported weather at BUF prior to the accident was the Automatic Terminal Information Service (ATIS) broadcast at 2154 EST. See Operations Group Chairman Factual Report, p. 3. The broadcast included the following information: wind 250 degrees at 15 knots gusting to 22 knots, visibility 3 miles in light snow and mist, with few clouds at 1,100’, ceiling 2,100’ broken clouds, 2,700 overcast, temperature 1, dew point -1, altimeter 29.79. Id. The ILS Runway 23 approach was in use at the time of the accident. See Operations Group Chairman Factual Report, p. 3. The Decision Altitude (DA) for the ILS Runway 23 approach is 928’ MSL (200’ height above touchdown) and the minimum required visibility is ½ statute mile. Id. At 2157, Cleveland Center cleared the flight to cross BENEE intersection at 11,000’ MSL. See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-43. At 2203, Cleveland Center instructed the flight to change to Buffalo Approach Control. Id. at 12-49. At 2203, Buffalo Approach Control advised that Buffalo altimeter was 29.80 and to plan an ILS approach for runway 23. Id. The flight crew confirmed the instructions and Captain Renslow briefed the approach at 2204. Id. at 12-50. At 2210, the first officer asked if there was ice on the windshield. See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-56. The captain responded that he had ice on his side, and asked “You don’t have yours?” Id. The first officer replied “Oh yeah oh it’s lots of ice.” Id. The captain said “Oh yeah that’s the most I’ve seen – most ice I’ve seen on the leading edges in a long time, in a while anyway I should say.” Id.

2

Buffalo Approach Control cleared the flight direct to TRAVA intersection and gave it a series of intermediate descent clearances, the last of which, at 2212, was to 2,300’ MSL. See Operation Group Chairman Factual Report, p. 3. According to ATC radar records, the flight crew intercepted the final approach course from the left. Id. Preliminary Flight Data Recorder (FDR) data show that at 2215:12 the crew extended the flaps from zero to 5 degrees. Id. At the time, the indicated airspeed was 172 kts., the autopilot was engaged and pitch trim was three degrees nose up. Id. At 2215:14, Buffalo Approach Control transmitted, “Colgan thirty four zero seven three miles from KLUMP turn left heading two six zero maintain two thousand three hundred til established localizer clear ILS approach runway two three.” See Operations Group Chairman Factual Report, p. 4. The flight acknowledged that clearance. Id. At 2216:02, when the airspeed was over 180 kts, preliminary FDR data showed the engine power levers were reduced to at or near idle and both engines’ torque values declined to minimum thrust. See Operations Group Chairman Factual Report, p. 4. The reduced power levers correspond to a sound heard on the CVR at 2215:59 similar to a decrease in engine power. See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-62. At 2216:07, Buffalo transmitted, “Colgan thirty four zero seven contact tower one two zero point five have a good night.” Id. At 2216:09, the crew extended the landing gear and the auto flight system captured the ILS 23 localizer. See Operations Group Chairman Factual Report, p. 4. Three seconds later the crew moved the engine condition levers forward to the 100% torque position, and at the same time, the crew radioed “Over to tower you do the same thirty four zero seven.” See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-62. At this time, airspeed had slowed to just over 160 kts. At 2216:23, the captain called for flaps 15 and the Before Landing Checklist. See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-62. At this time, airspeed was about 140 kts. See FDR Group Chairman Factual Report, p. 10-9. At 2216:24, either the left or right probes detected ice on the airframe, and three seconds later the flaps were moved to 10 degrees, the stall warning stick shaker activated, and the autopilot disconnected. Id. at 10-10, 10-14. The airspeed was 130 kts. Id. at 10-14. At this point, the captain pulled back on the yoke. Id. at 10-10. Less than two seconds following stick shaker activation, the control column went from 2 to 6 degrees, and the aircraft began to increase pitch at a rate of about 2.5 degrees per second (reaching a maximum of 30 degrees 6-7 seconds after stick shaker activation). Id. At 2216:31, the CVR recorded a sound similar to an increase in engine power. At 2216:34, the captain said “Jesus Christ.” See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 12-63. The crew added power to approximately 75% torque, and the captain again pulled back on the yoke, increasing control column pitch from 1.1 degrees to 7 degrees in less than one second. See Operations Group Chairman Factual

3

Report, p. 4. The airplane experienced a left roll, followed by a right roll, during which the stick pusher activated. Id. During this time, the indicated airspeed continued to decrease to less than 100 kts. Id. Eight seconds after the flaps had been selected to 10 degrees, and at an airspeed of less than 110 knots, the flaps began to retract. At 2216:37, the first officer said “I put the flaps up.” See Cockpit Voice Recorder Group Chairman Factual Report Addendum, p. 1263. Sixteen seconds later the flaps were fully retracted. At 2216:45, the first officer asked if she should put the gear up. Id. Following further pitch and roll excursions, the airplane pitched down and entered a steep descent from which it did not recover. See Operations Group Chairman Factual Report, p. 4. 1.2

Damage to Aircraft

The aircraft was totally destroyed due to impact forces and post-accident fire. The following statement is taken from the NTSB Structures Field Notes: The airplane was severely fragmented and a significant post crash fire destroyed most of the airplane and house. About 60% of the main structural components of the airplane were conclusively identified. Numerous small pieces of the airplane were recovered but not conclusively identified. Structure from the radome and both wingtips was identified in the wreckage. The empennage was intact in the wreckage. The airplane wreckage was on a heading of about 070º magnetic. Almost the entire forward fuselage was consumed in the post crash fire. The right, forward cargo door was recovered in the front yard of a house across the street (east) of the accident residence with impact damage but very little fire damage. A lower section of the left airstair door and the overhead cockpit emergency exit were recovered in the forward fuselage debris. Some pieces of the cockpit and windshield post structure was recovered on the northeast side of the house. A portion of the center fuselage and attached center wing box was recovered intact on top of the wreckage pile at the northeast corner of the house with severe fire damage. The recovered fuselage had collapsed vertically downward just below the cabin windows on both the left and right sides. Almost the entire aft fuselage was recovered in the wreckage. The portion aft of the aft entry doors did not exhibit fire damage. The aft right entry door remained installed and the aft left entry door and aft cargo door were recovered in the wreckage separated from the fuselage. The empennage was intact and attached to the aft fuselage. The empennage was rolled to the right and resting on the right horizontal stabilizer tip. The horizontal stabilizers, vertical stabilizer, elevators, and rudder were all intact and exhibited only slight impact damage to the leading

4

edges. The APU was in the tailcone but broken from its mounts. The left wing was entirely consumed by fire except for the aileron, outboard flap and a portion of the spoiler. These items exhibited severe fire damage and were recovered upside down, and rotated 180º from the direction of impact. The left engine mount and thrust structure was recovered in the wreckage with severe fire and impact damage. A small piece of the lower outboard wing skin was also recovered in the wreckage with severe fire and impact damage. The outboard right wing impacted the garage of the adjacent house. A section of the right wing outboard leading edge was found in the garage holding up the roof. Most of the outboard right wing structure including the engine support and thrust structure was recovered forward of the ground scar at the southeast corner of the house with no fire damage but had significant impact damage. A section of the inboard right flap was recovered and identified in the basement with moderate fire damage. Six propeller blades were recovered embedded in the ground at the ground scar location. All three landing gear were recovered in the wreckage. The right main landing exhibited minor fire damage to the tires and the retract actuator was broken off in the almost fully retracted position. The RMLG uplock was in the open position. The LMLG was recovered with severe fire damage and the retract actuator was broken off in the almost fully retracted position. The LMLG uplock was not conclusively identified. The NLG was recovered with severe fire damage. No conclusive evidence of the NLG position was found. Based on the tree strikes, impact marks, and ground scars, the aircraft’s pitch and roll angle at the time of impact was estimated to be approximately 30 degrees nose down and 30 degrees right wing low. 1.3

Other Damage (Ground Damage)

The following statement is taken from the NTSB Structures Field Notes: The airplane impacted a house located at 6038 Long Street, Clarence Center, New York. The house was a two story old style farmhouse with a basement, covered front porch, and detached two-car garage. The house and two cars were destroyed by the impact and post crash fire but the garage remained intact. The garage of an adjacent house to the south was also impacted by the airplane. The airplane impacted the south side of the house near ground level and pieces of the airplane traveled through the house and ended up beyond the northeast corner of the house foundation. The south basement wall was fragmented by the impact and most of the debris was pushed towards the northeast corner of the basement. The top of the north basement wall exhibited impact damage adjacent to the location of the left engine. There was a distinct ground scar on the south side of the house about 10 feet south of the south foundation wall. The ground scar was about 15 feet long, 5 feet wide, and 3 feet deep. Two trees along the southern property boundary were impacted and the tops were sheared off. The west tree was cut about 20 feet above ground level and the east tree

5

was cut about 25 feet above ground level. There was some tree debris between the property boundary and the house that exhibited clean, angled cuts consistent with propeller impact. 1.4

Personnel Information (a)

Flight Crew Information

Both flight crewmembers were certificated under Colgan Air and Federal Aviation Administration (FAA) certification requirements. See Operations Group Chairman’s Factual Report, pp. 4-7. According to FAA documents, there was no history of accidents, incidents, or enforcement actions against either the captain or first officer certificate numbers. See Operations Group Chairman Factual Report, pp. 6-7. They had no history of any driver’s license suspensions or revocations. Captain Renslow obtained his private pilot license in 1990 and his commercial pilot license in 2002. See Operations Group Chairman Factual Report, p. 4. He was hired by Colgan Air in September 2005, and held an airline transport pilot certificate (ATP) with type ratings in both the Saab 340 and Q400. Id. Captain Renslow’s background information is as follows: Date of birth: Date of hire with Colgan Air: ATP Certificate No. 480885562 issued: Airplane Multiengine Land Type Ratings: SF-340, Q400 (11-18-08) Commercial Privileges Airplane Single Engine Land Medical First Class issued:2

07-26-61 09-19-05 10-18-07

08-22-08

See Operations Group Chairman Factual Report, pp.4-5; Human Performance Group Chairman Factual Report, p. 3. Flight Times: Total flying time: Totat PIC time: Total Turbine time: Total time in type (Q400): Total flying time last 24 hours: Total flying time last 7 days: Total flying time last 30 days: Total flying time last 90 days: 2

3379 1030 3051 110.7 0 16:09 56:11 116:02

Limitation: “holder shall wear corrective lenses.”

6

Initial Type Rating (Q400):

11-18-08

Completed IOE (Q400): Last recurrent training: Last proficiency check: Last line check:

12-03-08 04-19-08 11-18-08 12-03-08

See Operations Group Chairman Factual Report, p. 6; Operations 2BB.

First Officer Rebecca Shaw obtained her private pilot’s license in December 2003 and her commercial license in September 2005. See Operations Group Chairman Factual Report, p. 7. She was hired by Colgan Air in January 2008. Id. at 6. Background information regarding FO Shaw is as follows: Date of birth: Date of hire with Colgan Air: Q400 SIC type rating (Certificate No. 2810129) issued: Commercial Pilot – Airplane Multiengine Land issued: Flight Instructor – Airplane Single Engine issued: Commercial Pilot- Airplane Single Engine Land issued: Medical First Class issued:

04-30-84 01-16-08 03-16-08 09-16-06 05-12-06 09-22-05 01-22-08

See Operations Group Chairman Factual Report, pp. 6-7; Human Performance Group Chairman Factual Report, p. 8. Total flying time: Totat PIC time: Total Turbine: Total time in type (Q400) Total flying time last 24 hours: Total flying time last 7 days: Total flying time last 30 days: Total flying time last 90 days:

2244 1251 774 774 0 15:49 57:20 163:21

SIC Q-400 Completed Initial Operating Experience: Last recurrent training: Last proficiency check:

03-16-08 03-22-08 01-15-09 03-16-08

See Operations Group Chairman Factual Report, pp. 6-7. (b)

Other Crew Members and Colgan Air Personnel

7

Flight Attendant Matilda Quintero, age 57, was hired in May 2008, and was trained and qualified to perform the duties assigned to her. Flight Attendant Donna Prisco, age 52, was hired in May 2008, and was trained and qualified to perform the duties assigned to her. Athough he was not a crew member of the accident aircraft, an off-duty Colgan pilot, Captain Joseph Zuffoletto, was on board the aircraft in the passenger cabin at the time of the accident. 1.5 Registration N200WQ

Aircraft Information Serial No. 4200

Line No. Date of Manu. A/C total time A/C total cycles 784 04-12-08 1819.3 1809

Type: Bombardier Dash 8-Q400 (DHC-8-402) Minimum Crew: two (pilot, co-pilot, and two flight attendants) The aircraft was operated and in compliance with 14 Code of the Federal Regulations (CFR) Part 121 and was in airworthy condition. 1.6

Meteorological Information

The last reported weather at BUF prior to the accident was the ATIS broadcast at 2154 EST, which stated: wind of 250 degrees at 15 knots, gusting to 22 knots, with visibility of three miles in light snow and mist, with a few clouds at 1,100’, ceiling at 2,100’ with broken clouds, and overcast at 2,700’. The temperature was 1 degree Celsius with a dew point of -1 degree Celsius, altimeter 29.79. The following information was taken from the Meteorology Factual Report: During the period of the KBUF upper air sounding or rawinsonde observation (RAOB), site number 72528, which was launched at approximately 1900 EST on February 12, 2009, a moist low-level environment with the lifted condensation level (LCL) 6 at 975-hPa or at 163 feet agl (868 feet msl), and a convective condensation level (CCL) 7 at 917-hPa or 1,757 feet agl (2,462 feet msl) was depicted. Surface temperature was 0.4 degrees C with a freezing level identified at 288 feet agl (933 feet msl). There were two frontal inversions identified at 5,703 feet msl and 10,460 feet msl with temperature remaining below freezing. The soundings moisture profile indicated a relative humidity of 75 percent or more from the surface to 16,000 feet msl, with relative humidity greater than 90 percent from the surface to 9,500 feet msl. The precipitable water content was 0.42 inches. The sounding temperature and moisture profiles supported

8

precipitation in the form of snow at the surface, and supported airframe icing conditions between the surface and 16,000 feet. The temperature inversions resulted in a Lifted Index (LI)8 of 11.8 and indicated a stable atmosphere and supported stratiform type clouds, including nimbostratus. The sounding wind profile indicated surface winds from 250 degrees at 20 knots gusting to 25 knots, with winds veering to the west-northwest with height through 10,000 feet. A low-level wind maximum was identified at 4,700 feet msl with winds from 297 degrees at 50 knots. The maximum wind was from 285 degrees at 73 knots at 38,700 feet, with the tropopause height at 29,172 feet. At 2,700 feet, the wind was identified from 265 degrees at 36 knots, with a temperature of –3 degrees C. Buffalo Niagara International Airport is a level B augmentation 12 airport with eight certified weather observers and is manned 24-hours a day. The augmenter’s office is located in an office on the ground floor of the terminal building. On the evening of February 13, the weather group chairman interviewed Rick Davis, the NWS Certified Weather Observer who worked the evening shift on the night of the accident and monitored him making several observations, and reviewing the weather observation logs on the night of the accident. Mr. Davis’s shift is from 1600 to 2400 daily Monday through Friday. His experience includes 9 years in the United States Air Force, with years as a weather observer and 6 years as a forecaster. He was born and raised in the Buffalo area, and has been an observer at KBUF for 8 years. On the night of the accident Mr. Davis indicated that the present weather was light snow, which consisted of large wet snow flakes. He indicated that at 2114 EST the ASOS system visibility sensor wanted to report 2 1/2 miles in light snow, but he could see the visibility markers at 3 miles and edited the system accordingly. He also edited the sky condition report at 2154 and 2249 EST based on his observation of the cloud coverage. The most significant event was at 0012 EST as he was briefing his relief and the ASOS issued a report of light freezing rain, Mr. Davis was not able to stop the issuance of the report and issued a correction immediately afterwards. Mr. Davis indicated that the large wet snow was still occurring and temperatures at the surface were above freezing, and that no mixed precipitation or icing was being observed. The same problem was also noted at Niagara Falls Airport (KIAG). The problem was thought to be in the software logic of the system. I viewed the observation logs and in fact a correction was noted, but it is not reflected in the observations in sections 2.0.1 and 2.0.2 of this report. The sensor issue will be addressed to NWS Headquarters. Mr. Davis also indicated the ASOS cannot detect freezing drizzle or ice pellets, and other mixed freezing precipitation in automated mode.

9

TAMDAR (Tropospheric Aircraft Meteorological Data Reports) are automated weather reports from commercial aircraft. These data are transmitted from the participating aircraft via ACARS. Two TAMDAR equipped aircraft reported icing between approximately 4,500 feet and 10,000 feet around the time of the accident. Figure 11 shows tracks of TAMDAR aircraft between 2100 EST (0200Z) and 2359 EST (0459Z). The two flight tracks to/from Buffalo both show icing (designated as red). Note that there is one TAMDAR flight flying well above Buffalo (from SE to NW) that is not reporting icing, but that plane was above the icing layer at around 16,000 feet. On the left is aircraft designated as #9392 that departed KBUF for KIAD at approximately 2137 EST (0237Z) or about 45 minutes prior to the accident. On the right is an aircraft designated as #9374 that operated from Washington Dulles International Airport (KIAD) to Buffalo (KBUF) arriving into KBUF approximately 2325 EST (0425Z) or approximately an hour after the accident. Figure 12 shows the raw data obtained from aircraft #9392 during its ascent from KBUF. Shortly after departure at 2139 EST (0239Z), the aircraft encountered icing at 6,000 feet with an air temperature of –8.5 degrees C and remained in icing conditions through 11,940 feet. The winds at 6,000 feet were from 304 degrees at 38 knots and at 11,940 feet, the aircraft reported winds from 309 degrees at 40 knots. Figure 14 shows the raw data obtained from aircraft #9374 during its flight from KIAD to KBUF with an arrival time of approximately 2325 EST (0425Z). At 2313 EST (0413Z) at 8,830 feet, the aircraft encountered icing conditions continuing through 4,590 feet. Winds at 8,830 feet were from 304 degrees at 37 knots while at 4,590 feet they were reported from 289 degrees at 42 knots. From 4,070 feet to 3,180 feet, the aircraft did not report icing conditions. The aircraft re-entered icing conditions at 3,120 feet and remained in icing through 2,360 feet. The aircraft only reported winds at the top of this icing level from 285 degrees at 37 knots. From 2,000 feet msl until the final data point, this aircraft did not report icing conditions. The Current Icing Product (CIP) is a supplementary weather product that provides a graphical view of the current icing environment. Input from weather sensors is provided to software models to produce this automatically generated graphical weather product. The CIP is updated hourly, and provides current information via icing severity graphics and icing probability graphics. CIP products are not forecasts, but presentations of current conditions at the time of the analysis. CIP is not to be used as a forecast for icing conditions. Figure 16 shows the CIP valid at 2200 EST (0300Z) for 3,000 feet msl. The chart indicates a greater than 80 percent chance of encountering icing at

10

3000 feet over much of western New York State. Figure 17 shows the CIP for the same period at 5,000 feet msl. This chart indicates a 70 to 80 percent chance of encountering icing at 5000 feet over the Buffalo area. Figure 18 shows the CIP for 7,000 feet msl. This chart indicates a 60 to 80 percent chance of encountering icing at 7,000 feet msl over the Buffalo area.

Aerodata observed that the flight which followed the accident flight 30 minutes later put in an “ICING” remark when they were just outside the marker on approach. See Operations Group Chairman Factual Report, Addendum 3, Attachment 5, p. 2. 1.7

Aids to Navigation

BUF Runway 23 has an FAA-certified ILS precision approach procedure that was fully operational at the time of the accident. All aids to navigation for BUF were operative at the time of the accident. 1.8

Communications

VHF communications were in use by the accident aircraft. All communication equipment for N200WQ, and ATC facilities for the route of flight, were fully operational on the day of the accident. 1.9

Flight Recorders

On February 13, 2009, the Safety Board’s Vehicle Recorder Division received the following FDR: Recorder Manufacturer/Model: Honeywell SSFDR, Model 980-4700, 128 Word Recorder Serial Number: 14241 The recorder was in good condition and the data were extracted normally from the recorder. 1.10

History of Colgan Air, Inc.3

Colgan Air, Inc. is a regional airline headquartered in Manassas, Virginia. It operates as Continental Connection, United Express and US Airways Express. Colgan Air offers more than 350 daily flights to 53 cities in 15 states and Canada. Colgan Air operates 34 Saab 340 and 14 Q400 regional airliners, and has hub operations in Boston, Houston, Newark, and northern Virginia. Colgan Air has approximately 1,300 employees. Colgan Air, Inc. was founded in 1991 and has gradually expanded its operations

3

This information may be found on page 25 of the Operations Group Chairman Factual Report.

11

over the years. In 1991, it began operating a flight with Beechcraft 1990C equipment between Washington Dulles International Airport and Binghamton, New York under the name National Capital. This route was later dropped and the name Colgan Air adapted. On July 1, 1997, Colgan Air began operating Continental Connection flights through a marketing alliance-code share agreement with Continental Airlines. On December 11, 1999, Colgan Air started operating under a code-share and service agreement as a US Airways Express Carrier. Colgan Air reintroduced Continental Connection service in March 2005 and acquired additional Saab 340 aircraft out of Houston. On October 4, 2005, Colgan Air started providing flights for United Express flights out of Washington’s Dulles International Airport. On January 17, 2007, Colgan Air was acquired by Pinnacle Airlines Corp., the parent company of Pinnacle Airlines, Inc., a regional airline doing business as Northwest Airlink. On February 2, 2007, Continental Airlines, Inc., Colgan Air and Pinnacle Airlines Corp. entered into a Capacity Purchase Agreement, which included code-share provisions, allowing Colgan Air-operated Q400 aircraft to do business as Continental Connection. On February 5, 2007, it was announced that Colgan Air would provide service for Continental Airlines, as Continental Connection, out of Newark Liberty International Airport starting in early 2008. Colgan acquired 15 Bombardier Q400 aircraft for that service as per the Capacity Purchase Agreement. 1.11

Organizational and Management Information

The organizational and management structure of Colgan Air pertinent to the issues set forth herein is as follows: The Vice President-Flight Operations is the executive manager in charge of operations. Reporting to him are the Director-Inflight, Director-Flight Standards, DirectorCrewmember and Dispatcher Training, Director-System Operations Control and Scheduling, Director-Flight Operations, and Director-Crewmember Resources. See Operations Group Chairman Factual Report, p. 26. Reporting to the Director-Flight Standards are the Manager-Inflight Standards, the Manager-Flight Standards, the Saab Fleet Manager, the Q400 Fleet Manager, and the APDs and Check Airmen. See Operations Group Chairman Factual Report, p. 26 Reporting to the Director-Crewmember and Dispatcher Training are the ManagerCrewmember and Dispatcher Training, the Manager-Inflight Training, and a clerk and an assistant. See Operations Group Chairman Factual Report, p. 27 Reporting to the Director-Operations is the Chief Pilot, who supervises regional chief pilots in EWR and IAH.4 See Operations Group Chairman Factual Report, p. 27.

4

Subsequent to the accident, a regional chief pilot was added in IAD.

12

1.12

Flight Schedule of Flight Crew

After finishing a trip on February 5, Captain Renslow had the next five days off work. On February 9, 2009, Captain Renslow spent the day at home in the Tampa area. He departed Tampa that afternoon at 1713 and arrived in EWR at about 2005. At 2247, he made a 6 ½ minute phone call. In post-accident interviews, the first officer who flew with him the next day stated the captain spent the night of February 9 in the EWR crew room. 5

On February 10, Captain Renslow began a two day trip. The report time was 0530. The trip ended at BUF at 1259. The duty day was 7:49 (hh:mm) which included 4:36 of flight time. Captain Renslow had the rest of the day off. According to the first officer who was flying with Captain Renslow that day, they spent the afternoon and evening relaxing at the hotel. The first officer last saw Captain Renslow about 2100-2130 as they left the poolcommunity area of the hotel to go to their rooms. Phone records indicate the captain made a 30 minute call at 2102. On February 11, the first officer reported seeing Captain Renslow in the hotel breakfast area at approximately 0500. After an almost 17 hour rest period, Captain Renslow began his duty day at 0615. His duty day was 9:49, which included five hours of flight time. His last flight on February 11 arrived at EWR at 1544. During post-accident interviews, the first officer said Captain Renslow had told him he was going back to his apartment that evening. The first officer characterized the flights on February 10 and 11 as unremarkable. The first officer stated the captain’s health appeared very good and the captain was well rested and alert during the trip. After his last flight on February 11, Captain Renslow did not fly again until the accident flight, over 27 hours later. On February 11, Captain Renslow placed or received calls between 1552 and 1637, 1823 to 1829, and 2020 to 2142. Crew Trac logon records indicate that he had multiple logons from 1610 to 1640, 1759 to 1831, and 2059 to 2110, that were coded under “access crew menu” and “self notify.”6 The last logon occurred at 2151. On February 12, at 0310, Captain Renslow logged onto the CrewTrac system, accessed the crew menu and acknowledged a revision to the accident trip’s schedule which showed a report time of 1330 and Flight 3407’s scheduled 1910 departure and 2048 arrival at BUF. Captain Renslow logged on again at 0726, and used his phone several times between 1000 and 1100. Between 1200-1400, he volunteered to perform some office work for the regional chief pilot at the EWR base operations office. Later in the afternoon, he was observed in the EWR crew room watching TV and talking to other pilots. At 1624, he made a call and at 1649 he received a call. Other calls were made in the early evening hours before the accident flight. 5

Information in this section regarding Captain Renslow was taken from the Human Performance Group Chairman Factual Report, pp. 6-7. 6 The access crew menu code indicates when a pilot logs onto the computer system to review schedules. The self notify code indicates the pilot is acknowledging a change in a trip.

13

First Officer Shaw had four days off duty preceding the accident flight.7 Her last flight prior to the accident flight ended in EWR at 1455 on February 8, 2009. On February 9, 2009, telephone records indicate outbound SMS messages were sent between 2152 and 2218 PST. According to her husband, FO Shaw awoke at home in Washington on February 10, 2009, between 0900-1000 PST. At 0926 PST, she placed a call lasting about five minutes, followed by several calls between 0931-0949 PST. She later went skiing in the Seattle area, returning home that afternoon. The evening was spent at home watching TV with her husband, who stated she went to sleep between 2000-2200 PST. At 2020 PST, she made a three minute call; and at 2214 PST she sent a SMS message. FO Shaw’s husband stated that she awoke on February 11, 2009, between 09001000 PST. At 0948 PST, she logged on to the CrewTrac system. At 1025 PST, a call was made to a number associated with FedEx jumpseat reservations. First Officer Shaw’s husband said she had reserved a jumpseat on a flight scheduled to depart SEA about 1900 PST, and she had to be at the airport one hour before departure. He drove FO Shaw to the airport (SEA) and, after stopping near the airport to eat, arrived about 1730 PST. FO Shaw traveled from SEA to MEM on FedEx Flight 1223. According to the captain of the flight, the flight departed about 43 minutes late. When the other passenger, a FedEx pilot, learned that the flight was not departing on time, he asked the first officer if the delay would affect her plans to get to EWR. She informed him she had a direct Continental Airlines flight as a backup. FO Shaw and the passenger rode in the cabin. During the first hour of the flight they talked about aviation and her career, and then he observed that the first officer went to sleep “solidly” for an hour and a half. The passenger said the lights in the back were dimmed. The flight arrived at MEM about 2330 PST. The crew and the two passengers rode the bus to the air operations center and the other passenger took the first officer to the jumpseat lounge to check-in. At 2356 PST, FO Shaw received a call on her cell phone that lasted three minutes. Between 2355-0040 PST, she sent eight text messages and received four. On February 12, FO Shaw traveled from MEM to EWR on FedEx Flight 1514. The flight departed MEM about 0418 EST8 and arrived at EWR about 0623 EST. The first officer rode in the cabin. After the flight landed, she told the captain she had slept the entire flight.

7

Information in this section was taken from the Human Performance Group Chairman Factual Report, pp. 10-13, and Human Performance Group Chairman Factual Report, Addendum 1, pp. 3-6. 8 The remainder of the times set forth in this section are in Eastern Standard Time.

14

At 0651, FO Shaw logged onto the CrewTrac system and acknowledged changes to the trip sequence. Another crew who had a 0655 report time and 0754 departure observed the first officer standing with the accident captain in the crew room. At 0729, phone records indicate a SMS message was sent. At 0732, the first office made a 2minute call to a number associated with Colgan Air operations. At 0829,9 the first officer sent the following text to her husband - “So I made it to Newark just fine and the weather is so [expletive] they cancelled my flights! I just go to Buffalo tonight at 730! I’m going to sleep I’ll call when I’m up.” A first officer reported seeing both pilots sleeping in the morning. A captain reported observing both pilots asleep around noon. A flight attendant reported the first officer sleeping in the late morning/early afternoon. After she awoke, the first officer sent the following text to her husband - “I feel soooo good, I took a nice 6 hour nap on the comfy recliner!” (According to text messages transcribed by Troy Shaw, this message was sent at 1405.) Phone records indicate SMS message activity for FO Shaw at 1305-1312, 1355, and 1416-1424. According to phone records, FO Shaw placed a call to her husband at 1425. Mr. Shaw confirmed that he spoke with her sometime in the afternoon between 1400-1500, and recalled that she sounded great and was just waiting to get on the airplane for Flight 3407. At 1459, the first officer logged onto the CrewTrac system to read messages. SMS message activity for FO Shaw occurred at 1514, 1522-1534, and 1611-1612. At 1625, she logged on to CrewTrac to access the crew menu. Further SMS message activity occurred at 1635, 1646-1647, 1659, and 1723. At 1726, FO Shaw received a call, and at 1747 and 1851 she placed calls. A first officer whose flight arrived about 1853 saw both Captain Renslow and FO Shaw walking down the jet bridge to the accident aircraft as he was exiting the airplane. 1.13

Captain Renslow’s Checking and Training Prior to Joining Colgan Air

On October 1, 1991, Captain Renslow was disapproved for his initial instrument airplane flight rating. See Operations Group Chairman Factual Report, p. 5. The aircraft was a PA-28-181 and the tasks disapproved were Pilot Operation 3 Task A and B; Pilot Operation 2, Task C; Partial Panel VOR Approach, NDB Approach and Holding. Id. He completed the rating on October 25, 1991. Id. This disapproval was disclosed by Captain Renslow on his employment application with Colgan Air. See Human Performance Group Chairman Factual Report, p. 3.

Air.

10

Captain Renslow trained at Gulfstream International Airlines prior to joining Colgan Some of this training is summarized below.

9

According to the time stamp on text messages transcribed by Troy Shaw (0529 PST). This training was disclosed in post-accident interviews conducted by the NTSB. See Operations Group Chairman Factual Report, Addendum 3. Captain Renslow’s training records were not disclosed to Colgan 10

15

On October 22, 1998, Captain Renslow underwent a line check. See Operations Group Chairman Factual Report, Addendum 3, Attachment 1. He received satisfactory ratings for the following inflight items: communication procedures, navigation, smoothness, and cockpit organization. Id. He also received satisfactory ratings for judgment, compliance with FARs and SOPs, descent planning, approach briefing and crew coordination, and approach procedures. Id. There were no unsats in this check, however, the check airman noted that Captain Renslow was a little behind the aircraft during approaches, and needed to work on callouts on approach. Id. On May 2, 2001, Captain Renslow received a satisfactory rating on the following inflight maneuvers: approaches to stalls – landing configuration, approaches to stalls – takeoff configuration, and approaches to stalls – clean configuration. See Operations Group Chairman Factual Report, Addendum 3, Attachment 1. On instrument procedures, he also received a satisfactory for an approach briefing. Id. On September 19, 2001, Captain Renslow received an unsatisfactory on the following maneuvers: approach to stall – landing configuration, and two engine ILS approach procedures. See Operations Group Chairman Factual Report, Addendum 3, Attachment 1. He received satisfactory ratings on approach to stall – clean configuration, approach to stall – takeoff configuration, and two engine ILS with flight director. Id. In the normal system operations portion, he received satisfactory ratings on anti-icing and deicing and stall warning/overspeed devices. Id. Regarding abnormal system operations, he received satisfactory ratings for stall warning heat vane failure and surface deice failure. Id. He also received satisfactories for judgment, situation awareness, crew coordination, and use of company procedures, profile, checklists. Id. For SIM period 3, he had glideslope deviations and was outside of altitude parameters. Id. In SIM session 4, the entire approach was unstable, and his airspeed was well below Vref during zero flap landing. Id. It was noted that he was often deviating from altitude. Id. In SIM session 6, Captain Renslow showed improvement maintaining aircraft attitude but was still deviating from altitude too often. Id. In SIM 7, his airspeed was more than 10 kts. below Vref +10. Id. The examiner wrote “Fly correct airspeed!” Id. In SIM 8, the examiner noted constant deviations up to full scale on glideslope, ½ to ¾ scale localizer. Id. Captain Renslow’s training records were not at all unusual in regards to his progression. See Operations Group Chairman Factual Report, Addendum 3, Attachment 1. The difficulty with stalls in sim 3 and 4 was corrected. Id. He performed at least three afterward that were satisfactory. Id. On May 14, 2002, Captain Renslow was disapproved for his initial commercial single engine land airplane flight certificate. See Operations Group Chairman Factual Report, p. 5. The aircraft was a C-177-RG and the tasks disapproved were takeoffs, landings, go-arounds, and performance maneuvers. Id. He completed the flight check for

Air when Captain Renslow was hired because Gulfstream did not employ Captain Renslow as a pilot, and thus the records were not covered by the Pilot Records Improvement Act (PRIA).

16

the certificate on June 25, 2002. Id. This disapproval was not disclosed by Captain Renslow on his employment application with Colgan Air. Once Captain Renslow had completed his training at Gulfstream and 250 hours of contracted flight time, he was not retained because Gulfstream did not need the manpower. See Operations Group Chairman Factual Report, Addendum 1, p. 5. Gulfstream noted, however, that Captain Renslow’s performance was above average. Id. at 6. In Summer 2007, Captain Renslow asked about coming back to Gulfstream, but did not pursue it. Id. He would have been hired based on his record. Id. On April 9, 2004, Captain Renslow was disapproved for his initial commercial multiengine land airplane flight certificate. See Operations Group Chairman Factual Report, p. 5. The aircraft was a PA-44-180. Id. He was notified that the entire flight portion of the practical exam would need to be re-accomplished. Id. He completed the flight check for the certificate on April 29, 2004. Id. This disapproval was not disclosed to Colgan Air. 1.14

FO Shaw’s Checking and Training Prior to Joining Colgan Air

FO Shaw trained at the KLM Flight Academy. See Operations Group Chairman Factual Report, Addendum 3, Attachment 3. While at KLM, according to its Training Manager, she performed 600-1000 stall recovery procedures. Id. On Jan. 23, 2007, she performed stall recoveries from a full stall in a clean configuration, approach to stall in clean configuration, approach to stall in approach configuration, and approach to stall in landing configuration. The KLM AOM (PA28-181) section on cruise/descent (Dec. 1, 2006) instructed that to recover from a stall in clean configuration, the pilot should maintain the pitch attitude, apply full power and right rudder, and not allow the nose of the aircraft to rise as a result of the increased power. See Operations Group Chairman Factual Report, Addendum 3, Attachment 3.Also, the pilot should ensure the flaps are up, check power, and accelerate without altitude loss. Id. The version of the KLM AOM (PA28-181) dated November 1, 2004, stated that to recover from a stall in a clean configuration, the pilot should bring the nose under the horizon, apply full power and right rudder without allowing the nose to rise. See Operations Group Chairman Factual Report, Addendum 3, Attachment 3. It also instructed that once the aircraft is un-stalled, the nose should be raised to initial climb attitude, and the pilot should climb back to the initial altitude. Id. Safe recovery prevails over altitude loss, which should be minimized. Id. Flaps are checked up, power checked, and the aircraft levels off. Id.

17

1.15

Colgan Air Hiring Process11

Colgan Air pilots undergo a rigorous, multi-tiered evaluation process before they are hired. In fact, about two-thirds of those who are initially contacted for an interview are not offered a job. See Operations 2H, Interview of Vice President of Administration Mary Finnigan, p. 46. The process starts with an online application through Airline Apps. See Operations Group Chairman Factual Report, p. 27. This application process covers flight time, accidents, incidents, violations, DUIs, failed checkrides, and other items bearing on the fitness of a pilot applicant. Id. Applicants who appear well qualified are contacted for a phone interview involving in-depth inquiries of the items covered by the online application and the pilot’s general flying background and qualifications. Id. Applicants who successfully pass the phone interviews are invited for an in-person interview. Id. At the in-person interview, the applicant takes a 50 question written test that parallels the FAA’s Airline Transport Pilot written exam. Id. Those who pass the written exam undergo an interview conducted by a pilot recruiter or the Manager of Recruiting and qualified line pilots. Id. Applicants who pass this phase of the interview process are evaluated by a check airman in a full motion simulator. Id. This full motion simulator evaluation is not required under Federal Airline Regulations. Pilots who successfully complete all steps in this process become training candidates. Id. As a training candidate, the pilot must pass all ground training, simulator training, checkrides, and PRIA (Pilot Records Improvement Act) records and background checks prior to commencing their initial operating experience. Id. Colgan Air’s internal hiring guideline at the time of the accident was 600 total hours with 100 hours of multi-engine time. Id. This guideline was not a rigid requirement; Colgan Air evaluates the total pilot and recognizes that certain experience is more valuable. Id. at 28. 1.16

Pilot Records Improvement Act of 1996 (PRIA)

All Part 121 air carriers are required to comply with the Pilot Records Improvement Act of 1996 (PRIA). PRIA provides that before a pilot may begin service with an air carrier, the hiring air carrier must request and receive certain information regarding the applicant pilot. The information includes records from the FAA, other air carriers (or other entities that employ pilots), and the National Driver Register. The records that must be provided by the FAA are: 1) current airman certificates with associated type ratings and limitations; 2) current airman medical certificate including any limitations; and 3) summaries of FAA legal enforcement actions resulting in a finding by the Administrator of a violation that was not subsequently overturned.

11

The hiring process detailed herein is set forth in the Operations Group Chairman Factual Report, pp. 2728.

18

Under PRIA, the records that must be provided by air carriers are any records that pertain to the pilot that are maintained under 14 CFR Part 121 (except records which do not pertain to the individual’s performance as a pilot), such as proficiency and route checks, training, and releases from employment. The reporting period is five years – records over five years must not be provided unless the information concerns a revocation or suspension of an airman certificate or motor vehicle license and the revocation or suspension is in effect on the date of the request. 1.17

Freedom of Information Act (FOIA)

The Advisory Circular regarding the Pilot Records Improvement Act (PRIA) in effect at the time Captain Renslow was hired on September 19, 2005, was AC 120-68C, issued on January 28, 2004. See Exhibit A, attached hereto. This Advisory Circular did not contain any reference to an air carrier obtaining Notices of Disapproval through FOIA, nor did it set forth any procedures in effect at the FAA for FOIA requests. Id. In fact, the Advisory Circular did not reference FOIA at all. Id. On November 7, 2007, AC 120-68D was issued and included the following new language: NOTE: a request with a signed consent by the pilot/applicant may be used to authorize the FAA to release records of Notices of Disapproval for flight checks for certificates and ratings to an air carrier making such a request. Air carrier representatives involved in the pre-employment screening process may find this additional information helpful in evaluating the pilot/applicant. These requests, however, are not an integral part of the standard PRIA request process. Consult the PRIA program manager for details. AC 120-68D, ¶ 3-8(a)(3). This FAA Advisory Circular was not in effect at the time Colgan Air hired Captain Renslow, who had been hired over two years before the Advisory Circular was issued by the FAA. In fact, at the time AC 120-68D was issued, there was no published guidance on requesting Airman Notices of Disapproval from the FAA. No such requests were made by air carriers between the issuance of the Advisory Circular and the time of the accident, nor were there established procedures in place at the FAA for handling such requests. 1.18

Hiring History of Captain Renslow

After Captain Renslow successfully completed Colgan Air’s hiring process detailed in Section 1.16, above, he was hired on September 19, 2005. See Operations Group Chairman Factual Report, p. 4. His online application was accepted through Airline Apps, he passed the phone and in-person interviews, he passed his simulator check, and the PRIA and background check did not reveal any disqualifying information. See NTSB Public Hearing Transcript, pp. 311-15; 347-48. After the accident, however, Colgan Air

19

learned that Captain Renslow was not truthful on his employment application. See NTSB Public Hearing Transcript, pp. 347-48. Colgan Air’s employment application contains the following question: “Have you ever failed any proficiency check, FAA check ride, IOE or line check?” See Human Performance Group Chairman Factual Report, p. 3. In response, Captain Renslow answered, “Yes”, and provided the following explanation, “FAA check ride for instrument rating. Id. I missed the NDB approach, received additional instruction, then repeated the approach and passed.” Id. Captain Renslow, however, had two other FAA certificate disapprovals he failed to disclose. See NTSB Public Hearing Transcript, pp. 347-48. He was initially disapproved for both his initial commercial single engine land airplane flight certificate (May 14, 2002) and his initial commercial multiengine land airplane flight certificate (April 9, 2004). See Operations Group Chairman Factual Report, p. 5. Colgan Air followed PRIA and the guidance published by the FAA on obtaining pilot records, but was unable to receive this information, because Captain Renslow was not employed as a pilot at the time of these failures, and thus PRIA did not cover the disapprovals. Id. As discussed above, at the time of Captain Renslow’s hiring, the FAA had not notified air carriers of any options under FOIA to receive additional information, and the FAA did not develop procedures for handling such requests until after the accident. Id. 1.19

Hiring History of FO Shaw

Like Captain Renslow, FO Shaw successfully completed Colgan Air’s hiring process detailed in Section 1.16, above. She was hired by Colgan Air on January 16, 2008. See Operations Group Chairman Factual Report, p. 6. Her online application was accepted through Airline Apps, she passed the phone and in-person interviews, passed the simulator check, and the PRIA and background check did not reveal any disqualifying information. FO Shaw had one FAA certificate disapproval, which she disclosed prior to being hired by Colgan Air. See Operations Group Chairman Factual Report, p. 7; NTSB Public Hearing Transcript, p. 349. The disapproval was for an initial flight instructor certificate, which she subsequently obtained. See Operations Group Chairman Factual Report, p. 7 1.20

Colgan Air Q400 Training Program

Training for the initial cadre of Colgan Air Q400 pilots was provided by FlightSafety Canada pursuant to the Purchase Agreement between Bombardier Inc. and Pinnacle Airlines Corp. See Operations Group Chairman Factual Report, p. 40. Colgan Air's Q400 training program, which was essentially the FlightSafety generic training program with added enhancements pertinent to Colgan's operations, was approved by the FAA in August 2007. See NTSB Public Hearing Transcript, pp. 228-230. Colgan Air’s training

20

program was used for the training of the initial cadre of Q400 pilots, although all training courses and simulator sessions were taught by FlightSafety instructors using FlightSafety simulators. By September 2008, Colgan Air was teaching ground training, although FlightSafety simulators were still being used. Colgan Air’s training uses state-of the-art equipment such as a full-motion simulator, a flight management system trainer, and a ground flight simulator. See NTSB Public Hearing Transcript, p. 158. Captain Renslow obtained his training under Colgan Air’s FAA approved program, using Colgan Air instructors and Flight Safety International’s simulator in St. Louis, Missouri. See NTSB Public Hearing Transcript, p. 228. He completed his ground training at Colgan Air’s facilities in Manassas, Virginia. Id. First Officer Shaw’s training was done by Flight Safety, and her checking was done by approved Flight Safety personnel.12 See NTSB Public Hearing Transcript, p. 230. Colgan Air’s basic indoctrination training lasts 40 hours. See Exhibit B, Crew Member and Dispatcher Training Program Manual, p. 4A-1, attached hereto. This course includes company policies and procedures, applicable provisions of the FARs, meteorological training, and other items. Id. at 4A1-4A5. In addition to basic indoctrination training, pilots undergo CRM training and winter operations training. Id. at 4A-61, 4A-644A69. The students then undergo 80 hours of systems training. Id. at 4A-33. At the completion of systems training, they take a written exam. Id. Before beginning simulator training, each student on the Q400 has to complete an online FMS training course, as well as three CPT Sessions. See NTSB Public Hearing Transcript, p. 261; Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 6. These CPT Sessions use a ground training device to go over flows, abnormal procedures, and use of automation. If they pass the written exam, they move on to seven simulator sessions, one checkride (the check ride will be a type ride in most cases if pilot is captain) and one Line Oriented Flight Training (LOFT) session (real time trip using simulator without interaction by instructor). See NTSB Public Hearing Transcript, pp. 209-10. The time in the simulator is about 14 days, so a new hire at this point would have undergone about six weeks of training. Id. at 247-48. After satisfactorily completing all required training, including an FAA regulatory check ride, a pilot must satisfy the operating experience requirements of FAR Part 121, before being considered fully qualified for revenue service. The pilot must fly with an FAAapproved check airman for all OE (Operating Experience) flights. See NTSB Public Hearing Transcript, pp. 210. FAA regulations require 20 hours of OE, and permit a 50% reduction in all OE hours for each landing. For example, by regulation, a pilot with 11 12

Under OpSpec A031, Colgan Air is authorized to make arrangements with specified training centers listed in order to conduct instruction and/or evaluations. See Operations Group Chairman Factual Report, p. 33. The Part 142 Training Centers that are permitted to conduct Initial, Transition, and Upgrade Training are Flight Safety International at St. Louis, MO, Seattle, WA, and Ontario, Canada. Id. at 33.

21

hours and 10 landings could complete OE. At Colgan Air, a pilot must complete the full 20 hours, regardless of the number of landings. Also, although there is no FAA requirement to do so, Colgan Air requires pilots to observe four flights from the jumpseat prior to OE. See Statement of Dan Morgan, Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, p. 6. Colgan Air OE hours range from 20 hours to 50 hours depending on the pilot training course (i.e., new hire, transition, or upgrade). After completing the OE, the pilot must pass a line check before operating as a fully qualified crew member. See NTSB Public Hearing Transcript, p. 210. Colgan Air’s pilots also undergo recurrent training once a year. See NTSB Public Hearing Transcript, p. 192. The ground portion of the recurrent training lasts at least three days. Id. at 248. In addition, captains receive a simulator proficiency check twice a year and first officers once a year. Id. at 192. These simulator sessions last four hours. `

1.21

Captain Renslow’s Checking and Training at Colgan Air

As a result of the training program set forth in Section 1.20 above, Captain Renslow had over 120 hours of Q400 training, four observation flights with experienced Q400 pilots, over 20 hours flying with a check airman observing him, and two successful check rides. The Colgan Air Crew Member and Dispatcher Training Program Manual sets out details of performance standards on page 2-2 and the following pages. It reads as follows: Flight crewmembers will be required to receive a satisfactory grade on all flight maneuvers, procedures and duties. A satisfactory grade is obtained ONLY when the student demonstrates the ability to operate the aircraft/simulator in the manner that shows he/she is obviously the master of the aircraft, and with successful outcome of each maneuver never in doubt on a scale of 1 through 4, as indicated below in the Simulator Grading Legend: Flight/Simulator Training Grading Legend: 1. 2. 3.

4.

Trainee understands maneuver and completes it successfully. No further training is necessary. Trainee understands maneuver and completes it successfully. Further training for retention is necessary. Trainee understands maneuver but is unable to complete the maneuver with the degree of accuracy necessary for flight check purposes. Further training is necessary and will be considered unsatisfactory performance. Trainee does not understand maneuver and is unable to complete it successfully. He/she needs further instruction and explanation prior to

22

further flight or simulator training and will be considered unsatisfactory performance. Additionally, all maneuvers must be accomplished in accordance with the applicable completion standards as set forth in this program. Unsatisfactory grades indicate the requirement of additional training. The standards set forth in the ATP Practical Test Standard as revised are the completion standards for all Proficiency Checks. See Operations Group Chairman Factual Report, pp. 33-34. The pilot must be removed from line operations until he/she satisfactorily completes the training or check. See Exhibit B, Crew Member and Dispatcher Training Program Manual, p. 2-3. In the event the pilot requires additional training following a proficiency check and the pilot is then found to be proficient, the check airman must document the items originally failed and generate a training form for the training conducted. The check airman may then certify the pilot’s proficiency as satisfactory if all maneuvers and procedures are performed satisfactorily. See Operations Group Chairman Factual Report, p. 35. In order to maintain objectivity, checking and evaluations are not be performed by the same individual who conducted the training. Id. Also, checking and evaluations are performed without interference or influence from company management. Id. On October 28, 2005, Captain Renslow was graded as TP (train to proficiency) on his initial proficiency check in the SF-340 as First Officer. See Operations 2BB. The task area identified was normal/abnormal procedures. Id. The grade identifies that one or more check items needed to be repeated, but that the overall performance was satisfactory. Id. On October 17, 2006, Captain Renslow received an unsatisfactory grade on his Recurrent Proficiency Training event at Colgan Air in the SF-340. Id. At the time, Captain Renslow was a first officer. Id. The unsatisfactory tasks were rejected takeoffs, general judgment, landings from a circling approach, oral exam, and non-precision approach. Id. According to company records, he attended recurrent training, and then completed requalification proficiency training on November 1, 2006. Id. On October 14, 2007, Captain Renslow received upgrade LOFT and sim training in the Saab 340. Id. On October 15, 2007, Captain Renslow was disapproved for his initial airline transport pilot certificate. Id. The aircraft was a Saab SF-340 and the task disapproved was approach and landing with powerplant failure – multiengine airplane. Id. After Captain Renslow failed his initial type certification on the Saab 340, the Director of Training informed the Chief Pilot of the failure. See Operations Group Chairman Factual Report, Addendum 3, Attachment 6. The upgrade proficiency check

23

was performed by a FAA Aircrew Program Designee. Id. The Chief Pilot spoke with the examiner and ascertained that Captain Renslow did not meet the standards for the singleengine missed approach procedure. Id. The instructor retrained Captain Renslow on the single-engine missed approach procedure, and had him conduct single engine approaches, go-arounds without the use of the autopilot, and single engine approaches without use of the flight director. Id. Captain Renslow met the standards, and the instructor signed him off for his checkride. Id. After this retraining, Captain Renslow was retested and completed his upgrade proficiency check and received his Airline Transport Pilot certificate.13 The upgrade proficiency check was performed by a FAA Aircrew Program Designee. Id. Captain Renslow also completed Colgan Air’s Captain Upgrade Curriculum, including an eight hour course on captain’s duties and responsibilities. See Human Performance Group Chairman Factual Report, p. 15. The intent of the course is to help the captain transition roles. Id. The course focuses on the captain’s duties, CRM, and situational awareness. Id. at 15-16. Although Colgan Air requires its upgrading captains to complete this course, it is not required for FAA Pilot-in-Command (PIC) qualification, and the FAA does not mandate captain leadership training. Id. at 16. As Chief Pilot, Mr. Honan was head of the Pilot Upgrade Review Board at Colgan Air. See Operations Group Chairman Factual Report, Addendum 3, Attachment 6. In this capacity, he reviewed the training records and interviewed crewmembers who were getting ready to upgrade or transition. Id. When Captain Renslow put in a bid to transition to the Q400, Chief Pilot Bill Honan spoke with Captain Renslow about the transition. Id. Mr. Honan reviewed Captain Renslow’s training records and noted he had three successful checking events since his initial failure for his type certificate on the Saab 340. Id. Mr. Honan informed Captain Renslow that he would have to successfully complete a proficiency check prior to transition. Id. Captain Renslow passed the check and was entered in the next transition class. Id. Captain Renslow successfully transitioned to the Q400 and passed his DHC-8 Type Rating. Id. From the time Captain Renslow successfully completed his Saab 340 upgrade proficiency check, until the time of the accident, he did not fail a single checking or training event. See Operations 2BB. He passed an upgrade line check performed by a Colgan Air checkairman on October 31, 2007, an upgrade operating experience observed by the FAA on October 31, 2007, a recurrent line check performed by a checkairman on September 26, 2008, a transition PC/type check given by an FAA Aircrew Program Designee (APD) on November 18, 2008, and a transition line check administered by a checkairman on December 3, 2008. Id. Captain Renslow also successfully completed recurrent proficiency training on April 19, 2008, transition ground school on October 31, 2008, and transition simulator training on November 17, 2008. See Operations 2BB. 13

Due to a clerical error, Colgan Air’s records reflect that Captain Renslow completed his upgrade proficiency check on October 15, whereas FAA records reflect that the ATP rating was received on October 18.

24

The following reflects the successful checking/training events following Captain Renslow’s failed October 2007 Saab 340 captain upgrade check: Training Events Retrained (Loft/Upgrade Sim/Add’l Sim)14 Recurrent Proficiency Training Q400 Transition Ground School Q400 Transition Sim Training

10-14-07/10-15-07 04-19-08 10-31-08 11-17-08

Checking Events Captain Upgrade PC15 (APD) Captain Upgrade Line Check (checkairman) Captain Upgrade Operating Experience (observed by FAA) Captain Recurrent Line Check (checkairman) Q400 Transition PC/Type Check (APD) Q400 Transition Line Check (checkairman)

10-15-07 10-31-07 10-31-07 09-26-08 11-18-08 12-03-08

See Operations 2BB. At the time of the accident, Captain Renslow had 3,379 total hours of flight experience and was Airline Transport Pilot rated, which is the highest level of pilot certification available. See Operations Group Chairman Factual Report, p. 6; Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 5. He held a “type certificate” in the Q400 issued by the FAA. He had 172 hours of formal training on the Q400 aircraft, qualifying him fully in accordance with all applicable Federal Aviation Regulations. 1.22

FO Shaw’s Checking and Training at Colgan Air

FO Shaw did not fail any checking events at Colgan Air. See Operations 2BB. She successfully completed her initial simulator training in March 2008, and passed her initial proficiency check in March 2008. Id. She passed initial winter operations training on January 28, 2008, and recurrent winter operations training on January 15, 2009. Id. At the time of the accident, FO Shaw had 2,244 total hours of flight experience, with 774 hours in the Q400. See Operations Group Chairman Factual Report, p. 7.

14 15

The records of Colgan Air with respect to the dates of this retraining do not match the FAA’s dates. FAA records reflect this occurred on October 18, 2007.

25

1.23

Colgan Air Stall Warning Training

All Colgan Air Q400 pilots receive training in the Q400 Stall Protection System, including the stick shaker and stick pusher functions, stall recognition, and stall recovery, in their academic training during ground school. See NTSB Public Hearing Transcript, p. 170. At the time of the accident, pilots also received hands-on stall warning, recognition and recovery training (except for stick pusher demonstration) in the Q400 simulator. See Operations Group Chairman’s Factual Report, pp. 35-36; NTSB Public Hearing Transcript, p. 162. Like all other Part 121 carriers, Colgan Air’s FAA-approved training program teaches and checks “Approach to Stalls.” See Operations Group Chairman Factual Report, p. 35. Pilots learn how to recognize an impending stall and how to recover from it before an actual aerodynamic wing stall occurs. See NTSB Public Hearing Transcript, p. 162. The stick shaker is designed to warn a pilot about an impending aerodynamic wing stall. Id. at 35, 50. The stick pusher activates after the stick shaker and only after the Stall Protection System has determined that an aerodynamic wing stall has occurred. Id. at 4950. Stick pusher is not a warning function; it is a post-stall recovery tool. Id. at 50. All Colgan Air pilots are trained to respond to a stick shaker so that an actual aerodynamic wing stall is avoided. If the crew performs the proper approach to stall recovery procedure, they will never experience an aerodynamic stall or stick pusher. Stick pusher simulator demonstration training is not a standard practice in the airline industry. The Colgan Air DHC-8 Q400 Series Simulation Training is described in Appendix F in the Training Manual, pp. F-1 to F-13. See Operations Group Chairman Factual Report, p. 35. The training described is for Q400 initial, transition, upgrade, recurrent16 and requalification pilot simulator training. Id. Flight training events are described in detail for eight flight simulator training modules, the last of which is a check ride. Id. The Q400 training syllabus provides instruction on “approach to stalls” in three different flight configurations – takeoff, landing, and clean. See Operations Group Chairman Factual Report, pp. 35-36. These sessions are trained in three separate fourhour modules of the simulator training course, with each pilot in the two-pilot crew encountering stall scenarios. The student will spend as much time training on these specific maneuvers as necessary to perform them at the highest level of proficiency. During their certification check ride, all students are tested on stall recognition and recovery in all three flight configurations. Stall warning in the Q400 simulator training is conducted in lessons 1, 4 and 7 and stalls are evaluated on the proficiency check. See Operations Group Chairman Factual Report, p. 35. The following training for stalls is programmed:

16

Recurrent training (ground portion) is 3 or 3.5 days. Recurrent training (sim portion) is 8 hours that includes a four hour simulator session, an oral exam, prior briefing, and de-briefing.

26

Lesson One – Approach to Stall – takeoff, enroute, and landing configuration Lesson Four – Approach to Stall – takeoff and landing configuration Lesson Seven – Approach to Stall – landing configuration Id. Illustrations of the three stall profiles to be flown are depicted in the CFM, section ten, revision 1, pp. 7-9. Id. The clean stall, which refers to a stall performed with gear and flaps retracted, is entered from an airspeed of 180 kts, at a minimum altitude of 5000 feet AGL, and with power at flight idle. See Operations Group Chairman Factual Report, p. 35. The pilot flying (PF) calls out “stall,” advances power to the rating detent, and states “check power.” Id. The profile says that during the stall the PF is to maintain heading and altitude. Power should be adjusted to maintain 180 kts to exit the maneuver. Id. The takeoff stall is entered at 180 kts, at a minimum altitude of 5000 feet AGL, with flaps set to 15 degrees, gear down, and power at flight idle. See Operations Group Chairman Factual Report, p. 35. The PF is to maintain heading and altitude during the maneuver and begin a 20 degree bank turn at 120 kts. Id. The PF calls out “stall,” advances power to the rating detent, rolls wings level, and states “check power.” Id. The PM calls out “positive rate,” the PF calls “gear up,” the PM calls “Vfri,” and the PF calls “flaps 0.” Id. Power should be adjusted to maintain 180 kts to exit the maneuver. Id. at 36. The landing stall is entered at 180 kts, at a minimum altitude of 5000’ AGL, with flaps set to 35 degrees, gear down, and power at flight idle. See Operations Group Chairman Factual Report, p. 36. The PF is to maintain heading and altitude during the maneuver. Id. The PF calls out “stall,” advances power to the rating detent, and states “check power, flaps 15.” Id. The pilot monitoring (PM) calls out “positive rate,” the PF calls “gear up,” the PM calls “Vfri,” and the PF calls “flaps 0.” Id. Power should be adjusted to maintain 180 kts to exit the maneuver. Id. The PTS standard is to recover from the stall with a minimal loss of altitude. See Operations Group Chairman Factual Report, pp. 36-37. During the stall recovery exercises for initial simulator training, the candidates were instructed to maintain an assigned altitude and complete the recovery procedures with minimal altitude loss. The basic pilot proficiency check is detailed on pages 6-2 to 6-5. See Operations Group Chairman Factual Report, p. 36. Both PIC’s and SIC’s must perform approaches to stalls. Id. Regarding approaches to stalls (pp. 6-4), the manual states: Check airmen shall evaluate the applicant’s ability to recognize and recover from an approach to stall in three separate airplane configurations. The three configurations are the clean configuration, the takeoff configuration, and the landing configuration. Check airmen may waive all but one of the stalls and the one stall must be performed while in a turn with a bank angle between 15 and 30 degrees. This waiver authority should be used when an

27

applicant’s performance in other events indicates a high degree of proficiency. Id. For checkrides, the crew member prepares for the stall by setting the configurations appropriate to the stall, such as configuring the flaps, the gear, the condition levers, preparing for the stall, and announcing the area is clear before performing the stall. See NTSB Public Hearing Transcript, pp. 240-41. The stall maneuver itself requires proper company call-outs, and execution of procedure, followed by return to the flight regime. Id. The stall maneuver is done as a memory item. Id. 1.24

Flight Crew’s Stall Warning Training

Captain Renslow and First Officer Shaw received the stall training that was part of Colgan Air’s FAA approved training program for the Q400. See NTSB Public Hearing Transcript, pp. 161-62. Colgan Air meets all FAA requirements for stall training, and Colgan Air’s stall recovery procedures follow the FAA’s Practical Test Standards. See Operations Group Chairman Factual Report, p. 33. In addition, to obtain each grade of license, pilots must pass a certification check ride that includes stalls as a required maneuver (private pilot, instrument rating, commercial pilot, and multi-engine). Since stall recovery procedures are essentially the same for every aircraft – decrease the angle of attack, add power, and level the wings, the stall checking and training a pilot receives over his/her career has a cumulative effect. See NTSB Public Hearing Transcript, pp. 191, 194. During their Q400 check rides, both Captain Renslow and First Officer Shaw were tested on stall recognition and recovery. See NTSB Public Hearing Transcript, pp. 161-62. This would have been the sixth certification event at a Part 121 airline wherein Captain Renslow would have been tested on stall recognition and recovery. See Operations 2BB; Operation Group Chairman Factual Report, Addendum 3, Attachment 1. Prior to entering a Part 121 training program, every single pilot must have shown proficiency in recovering from a stall during four certification events – private pilot, instrument rating, commercial pilot, and multi-engine rating. Captain Renslow received stall warning checking and/or hands-on training at Colgan Air on the following occasions: October 6, 2005 October 28, 2005 October 17, 2006 November 1, 2006 October 3, 2007 October 14-15, 2007 November 17, 2008 November 18, 2008

Initial Simulator Training Initial Proficiency Check Recurrent Training Recurrent Training Upgrade Proficiency Check Simulator Training and Upgrade Proficiency Check Transition Simulator Training Transition Proficiency Check

28

See Operations 2BB. FO Shaw received stall warning checking and/or training at Colgan Air on the following occasions: March 11, 2008 March 16, 2008

Initial Simulator Training Initial Proficiency Check

See Operations 2BB. Captain Renslow performed at least 35 stall recovery procedures during his career at Colgan Air. Also, since the stick shaker is a first flight of the day test item, Captain Renslow felt the operation of the stick shaker each time he flew the Q400 on an origination flight. Review of company records show Captain Renslow flew the Q400 on 25 origination flights. See Spreadsheet of Captain Renslow’s Originating Flights in Saab 340 and Q400, attached to December 4, 2009 letter to Lorenda Ward. In addition to his training at Colgan Air, Captain Renslow was also trained in stall recognition and recovery at Gulfstream. For example, he received stall training in May 2001 and September 2001, which is summarized in Section 1.13, above. The Q400 training program did not include tactile stick pusher training. Instead, pilots were taught the proper approach to stall recovery techniques in response to the stick shaker. The training ensures an actual aerodynamic stall does not occur, so the stick pusher will not fire. Colgan Air’s Saab 340 training, however, did include hands-on stick pusher training. Captain Renslow underwent stick pusher training in the Saab 340 on the following occasions: October 6, 2005 October 14-15, 2007

simulator initial training upgrade simulator training

See Captain Renslow’s Training Record, Operations 2BB. The syllabus for lesson one of the Saab 340 simulator training for initial, transition, and upgrade training is a four hour session that includes a demonstration of the stick shaker and stick pusher in a clean configuration, takeoff configuration, and landing configuration. See Crewmember and Dispatcher Training Program, Saab 340 B Series Syllabus. Lesson one also contains a pre-flight stall warning test that includes an activation of the stick shaker and stick pusher which provides the pilot with hands-on feedback. Id. Captain Renslow also would have felt the operation of the stick pusher on the ground during the following training:

29

October 28, 2005 October 15, 2007 October 18, 2007 April 19, 2008

initial proficiency check failed type ride completed upgrade proficiency check and airline transport pilot certificate recurrent proficiency training

See Captain Renslow’s Training Record, Operations 2BB. In addition, Captain Renslow would have felt the operation of the stick pusher each time he flew a Saab 340 aircraft on a first flight of the day, since a stick pusher test was one of the “first flight of the day” check items. See December 4, 2009 letter to Lorenda Ward. The test involves holding the stall one and two switches in the upper position, and confirming that: 1) the left and right stick shakers activate and “clackers” sound; 2) both PUSH 1 and PUSH 2 lights illuminate; and 3) the stick pusher activates. Id. Review of company records reveals that Captain Renslow operated the Saab 340 on its origination flight on 89 occasions. Id. In sum, Captain Renslow had multiple training sessions on a stick pusher in the simulator, and also felt the operation of a stick pusher on almost 100 occasions. Id. This training is significant due to the similarities between the stall warning systems in the Saab 340 and the Q400. Id. The stall warning system for the Saab 340 and the Bombardier Q400 are nearly identical in all significant ways. Id. However, the Saab 340 does not have a switch like the Q400 REF SPEEDS switch that artificially raises the stall warning activation speeds. Id. Both include a stick shaker and a stick pusher, and the components and operation of each is essentially the same. Id. Both aircraft use two independent computers that receive angle of attack signals from two transducers located on either side of the fuselage. Id. If certain thresholds are reached, the stick shaker will activate and an aural warning will be provided. Id. If the angle of attack further increases, the stick pusher is activated. Id. Both the Saab 340 and Q400 stall warning systems have the same basic type of components, including the following: stall protection modules, stick pusher actuator, stick pusher capstan, stick pusher quadrant, a stick shaker, and angle of attack transducer. Id. The Saab 340 stall warning system is comprised of two stall warning identification computers, two angle of attack transmitters, two stick shakers (one on each control column), an aural stall warning, a visual warning system, and a stick pusher mechanism that transfers rotation of the stick pusher servo actuator to a forward movement of the control columns. Id. The pusher applies 80 pounds of force forward on the control column to a position of 4 degrees elevator down. Id. The pusher can be overridden by approximately 90 pounds of force exerted on the control column. Id. Similarly, the Q400’s stall protection system consists of two stall protection modules. Id. Each module uses the following parameters: angle of attack, flap position, mach number, power level angle, condition level angle, and REF SPEEDS switch position.

30

(See Section 1.27 for more details on the REF SPEEDS switch). Id. The modules calculate when to activate the stick shaker and stick pusher. Id. When the modules operate the stick shaker, a signal to the Automatic Flight Control System disengages the autopilot. Id. If the aircraft is near a stall condition, the control columns will vibrate. Id. In addition to this tactile warning, the stick shaker motor and the rattling of the chain mechanism on the control column creates a loud noise. Id. If the stall is not averted, the Q400’s stick pusher activates and provides a 70 pound nose-down force to both control columns. Id. The stick pusher actuator turns an output pinion to move a stick pusher capstan that moves the control column forward when the aircraft is in a stall condition. Id. The stick pusher requires 80 lbs. of force to override. Id. Due to the fact that the Saab 340’s stall warning system (including the stick pusher) is nearly identical in all functional ways to that of the Q400 (except for the REF SPEEDS switch), the operation of the Q400 stick pusher would have been quite familiar to Captain Renslow. 1.25

Colgan Air/Q400 Winter Operations Procedures

The Bombardier Q400 Aeroplane Flight Manual (AFM) provides procedures for the use of the Ice Protection Panel. See Operations Group Chairman Factual Report, pp. 1416. The AFM states the REF SPEEDS switch should be set to INCR prior to encountering icing conditions, when ice is first detected, or when the flashing “Ice Detected” advisory appears on the Engine Display. Id. at 16. This selection is to be done at the same time as the engine intake door switches, the prop selector, and the windshield heat are placed on; the pilot’s side window heat switches are to be turned on if ice is forming on the pilot’s side window. Id. For climb, cruise, and descent in icing conditions (paragraph 4.7.2.3), the AFM procedure is to maintain at least minimum airspeed (climb, Vfto + 20 kts; descent 1.223 Vsr for 0 flaps + 25 kts); to operate the de-ice boots on FAST or SLOW depending on rate of accumulation (SLOW, however, can only be used during cruise flight), and to monitor wing and tail advisory lights for normal operation. See Operations 2KK. When the aircraft is aerodynamically clean, the procedure is to turn off the de-ice boots and the REF SPEEDS switch. Id. At this time, the pilot should revert to non-ice airspeed. Id. A note states “The aircraft is aerodynamically clean when all ice is removed from the visible leading edges and wing tips.” Id. For holding, approach and landing in icing conditions (paragraph 4.7.2.4), minimum airspeeds in icing apply, the de-ice boots must be operated in FAST, and a performance penalty must be applied. Id. There is a note which states, “When holding in icing conditions flaps must be at 0 degrees.” Id. Minimum airspeeds for operating in icing conditions are as follows: -Minimum holding, 190 kt.

31

-Approach Speed 1.23 VSR (Figure 5 – 1 – 2) flap 0° + 25 kt. -Approach Speed (Figure 5 – 8 – 1) flap 5° + 20 kt. -Go-around Speed (Figure 5 – 8 – 1) flap 5° + 20 kt. -Approach Speed (Figure 5 – 8 – 2) flap 10° + 20 kt. -Go-around Speed (Figure 5 – 8 – 2) flap 10° + 20 kt. -Approach Speed (Figure 5 – 8 – 3) flap 15° + 20 kt. -Go-around Speed (Figure 5 – 8 – 3) flap 15° + 20 kt. -Landing VREF (Figure 5 – 8 – 1) flap 10° + 20 kt. -Landing VREF (Figure 5 – 8 – 2) flap 15° + 20 kt. -Landing VREF (Figure 5 – 8 – 3) flap 35° + 15 kt. Id. When the aircraft is no longer in icing conditions (paragraph 4.7.2.5), the procedure is to continue the use of de-ice boots on “fast” until all ice is removed from the visible leading edges. See Operations Group Chairman Factual Report, p. 16. Once the aircraft is aerodynamically clean, the procedure is to turn the de-ice boots and the REF SPEEDS switch off and resume normal airspeeds. Id. The Colgan Air CFM, section 2, Limitations, states under paragraph 2.6.6, Ice Protection: Engine: Engine intake by-pass doors must be open for engine operation in icing conditions. NOTE: Icing conditions exist when the SAT on the ground and for takeoff is 10 degrees C or below, or SAT in flight is 5 degrees C or below, and visible moisture is present in any form (such as clouds, fog with visibility of one mile or less, rain, snow, sleet or ice crystals. Icing conditions also exist when the SAT on the ground and for takeoff is 10 degrees C or below when operating on ramps, taxiways or runways where surface snow, standing water, or slush may be ingested by the engines or freeze on engines, nacelles or engine sensor probes. Airframe: When ice is detected, the AIRFRAME MODE SELECT selector must be positioned at FAST or SLOW. See paragraph 4.7, OPERATION IN ICING CONDITIONS. See Operations Group Chairman Factual Report, p. 16. The Q400 Engine Display (ED) shows a message, “ICE DETECTED,” when one or both ice detector probes have detected more than 0.5 mm of ice. See Operations Group Chairman Factual Report, p. 17. This is depicted by item one of Figure 7. Id. The message will flash in yellow reverse video for five seconds, and if the REF SPEEDS switch is not set to the INCR position, the message continues in normal video. Id. When the airplane is in icing conditions with the REF SPEEDS switch set to INCR, the message is in white normal video, not flashing. Id. Item two of figure 7 depicts the “INCR REF SPEED message, which displays when the REF SPEEDS switch is set to INCR. Id.

32

Colgan Air’s ground training curriculum includes a winter operations section specifically focused on how to handle winter weather conditions such as icing. See Operations Group Chairman Factual Report, p. 38. Winter operations training covers:     

use of holdover times; aircraft deicing/anti-icing procedures, checks and responsibilities; aircraft surface contamination, critical area identification and effect on performance and handling characteristics; types, purpose, characteristics, and effectiveness of deicing and antiicing fluids; deicing/anti-icing fluids handling/performance implications.

See Operations Group Chairman Factual Report, p. 38.

According to Colgan Air training records, Captain Renslow completed transition winter operations anti-ice/deice training on October 31, 2008, during his transition training for the Q400. See Operations 2BB. Training records also showed that FO Shaw had last completed recurrent winter operations anti-ice/de-ice training on January 15, 2009. Id. During winter operations training, as well as during initial and recurrent ground school, Colgan shows its pilots a video entitled “Icing for Regional & Corporate Pilots.” The video discusses the latest information from NASA and the FAA about icing. The purpose of the video, as stated in its introduction, is:     

To review fundamentals of aircraft icing To enhance the pilot’s ability to assess hazardous icing conditions To enhance the pilot’s understanding of icing effects on stability and control of the aircraft To present strategies that pilots can use to exit a hazardous icing encounter To discuss super cooled large droplets

See Operations Group Chairman Factual Report, p. 38. Initial ground school winter operations training is a two hour course; recurrent is one hour. See Colgan Air Crew Member and Dispatcher Training Program Manual, p. 4A-64. In the simulator for new hires, upgrade, and transition, the pilot will receive three to six hours of exposure to icing conditions, failure of deicing equipment, etc. Use of icing equipment is also covered in recurrent proficiency checks and in the oral exam of the proficiency check. Colgan pilots also receive training regarding deicing procedures and operation of deicing equipment when they undergo aircraft-specific training.

33

1.26

Q400 Airspeed Indication, Speed Bug and Low Speed Cue

The primary airspeed indications are on the left and right Primary Flight Displays (PFD). The Q400 AOM Section 12.12 describes the aircraft airspeed indication systems. See Operations Group Chairman Factual Report, Figure 9, p. 20. Indicated airspeed is displayed on the upper left side of the instrument. Id. The vertical scale shows the aircraft’s current indicated airspeed as a rolling drum indication with marks every ten knots. Id. The tape displays +/- 42 KIAS around the actual aircraft speed. Id. A digital display shows the IAS from 30 KIAS up to 500 KIAS. Id. The number seven (7) shows the current airspeed. Id. The hollow triangle is the desired (bugged) airspeed, and the low speed cue is depicted by the number five (5). Id. The vertical bar below the low speed cue is red. Id. When IAS is less than or equal to the low speed cue, the digital IAS displays changes to red. Id. The low speed cue is computed by the Stall Protection System and provides indication of minimum operating speeds. Id. 1.27

The Q400 Reference Speeds Switch

The Q400 has a reference speeds (REF SPEEDS) switch that the pilot sets to INCR when operating in icing conditions. According to the Bombardier AOM, when the REF SPEEDS switch is set to INCR, the stall firing angle of the stick shaker and the stick pusher is set to a relatively lower angle of attack, which means that the Stall Protection System changes its calculations because of the reduced performance limitations of the aircraft in icing conditions. This has the effect of increasing the speed at which the stall warning system (i.e., the stick shaker and stick pusher) will activate, providing the same stall warning margins whether or not operating in icing conditions. When the REF SPEEDS switch is set to INCR, the Engine Display (ED) of the Electronic Instrument System (EIS) shows a white INCR REF SPEED message. No red or amber Warning or Caution lights are illuminated when the REF SPEEDS switch is set to INCR. The Q400 is the first variation of the Dash 8 that incorporates the use of the REF SPEEDS switch to achieve the required stall warning margin for icing operations. In order to maintain the same margin between stall warning activation (stick shaker) and actual stall speed for operations in icing conditions, when the REF SPEEDS switch is set to INCR, the flight crew is required to add an additional 15, 20, or 25 kts to the normal Vref speed depending on the expected landing flap setting. Assuming a flaps 15 landing with the REF SPEEDS switch set to INCR, the AFM and AOM require that the Vref speed must be increased by 20 kts. There are no “Cautions” or “Warnings” in the AFM or AOM that setting a Speed Bug to a nonicing Vref speed with the REF SPEEDS switch set to INCR could result in stick shaker activation at a speed greater than Vref. However, there is a “Caution” in the AFM and AOM in the “Take-Off In or Into Icing Conditions” that states “If airspeed is

34

not increased before the REF SPEED switch is set to INCR, stall warning may occur.” There is nothing to prevent a pilot from inputting a Speed Bug speed that is below the Stall Warning System’s calculated stick shaker activation speed. Further, there is no active warning to a pilot that a bug speed has been set below the calculated stick shaker speed or Low Speed Cue. Eight days after the accident, Colgan Air issued an Operations Bulletin which reinforced the operation and use of the REF SPEEDS switch and included the following warnings and cautions: CAUTION: If airspeed is within 20 Knots of the Low Speed Cue, the airspeed must be increased before REF SPEEDS switch is selected to INCR or a stall warning may occur. CAUTION: If Vref [not Vref (ICE)] is used for landing, the REF SPEEDS switch must be selected to OFF or a stall warning may occur at a speed higher than Vref. See Operations 2LL.

On March 18, 2009, Colgan Air issued Operations Bulletin Q400 CFM #09003, which added a REF SPEEDS switch item to the Normal Checklist. See Operations 2MM. The position of the REF SPEEDS switch is to be checked by the crew after the Approach & Landing Brief is completed and before the speed bugs are set. The bulletin also states as follows: The decision to turn the REF SPEEDS switch to INCR or OFF, for an approach, should be made prior to entering the initial approach phase and before the final approach segment. It is prohibited to change the position of the REF SPEEDS Switch below 1,000’ AGL Whenever the REF SPEEDS switch is set to INCR the only speeds which may be bugged are Vref (Ice) and Vga (Ice). When the REF SPEEDS switch is set to OFF the speed bug may be changed to Vref and Vga if above 1000’ AGL. Id. In addition, the Adverse Weather section of Colgan Air’s current FAAapproved Q400 Company Flight Manual now includes the following guidance, warnings and cautions regarding the REF SPEEDS switch:

35

CAUTION: If airspeed is within 20 Knots of the Low Speed Cue, the airspeed must be increased before REF SPEEDS switch is selected to INCR or a stall warning may occur. CAUTION: If Vref [not Vref (ICE)] is used for landing, the REF SPEEDS switch must be selected to OFF or a stall warning may occur at a speed higher than Vref. CAUTION: If airspeed is not increased before REF SPEEDS switch is selected to INCR, stick shaker(s) may activate. See Exhibit C, attached hereto. 1.28

AeroData Calculations for Proper Landing Speed and Flight Crew Deviations

Colgan Air uses AeroData to provide takeoff and landing performance and weight and balance calculations. See Operations Group Chairman Factual Report, p. 14. Flight crews can request and receive real time calculations through ACARS. Id. After the crew enters required items, such as temperature, altimeter setting, runway, and airplane gross weight, the performance data is calculated and received within 10 to 30 seconds. Id. For landing data, optional keyword entries are “icing” and “eice.” Id. The system then returns a message which displays Vref, Vga, Vfri, and Vcl17, as well as maximum allowable landing weight and landing distance. Id. The icing entry is made if the flight is in icing conditions. The eice entry is made if ice is picked up during flight and residual ice stays on the airframe even after operation of the deice boots. Per the Bombardier AFM, the chart below shows the minimum reference speeds for icing conditions based on an estimated landing weight of 54,366 lbs. See Operations Group Chairman Factual Report, p. 8. The estimated landing weight computed in accordance with Ops Spec A099 was 54,306. Id. at 9. The landing weight entered by the crew into ACARS was 54,700. Id. at 8. Configuration Zero flaps Flaps 5

Clean 145 kts 133 kts

Icing 170 kts 153 kts.

17

Actual (Flight 3407) 17018 170 down to sub-14019

Landing reference speed, go-around speed, flap retract speed, and climb speed. Airspeed immediately before Flaps 5. 19 The decrease in air speed occurred over 74 seconds, but was not linear. Instead, air speed initially rose, then steadily declined from over 180 kts to under 140 kts in 26 seconds. 18

36

Flaps 10 Flaps 15

124 kts 118 kts

144 kts. 138 kts

130 n/a

See Operations Group Chairman Factual Report, pp. 10-13. Based on the crew’s Aerodata request, the Vref speed provided to the crew was 118 kts (flaps 15). See Operations Group Chairman Factual Report, p. 13. According to the CVR transcript, it appears that both flight crewmembers set 118 kts as Vref in the Speed Bug #1, despite the REF SPEEDS switch being set to INCR. At 2216, power was reduced to near flight idle and airspeed slowed from about 180 to 130 kts and the stick shaker activated about the same time the flaps were moved to ten degrees. (The stick shaker would not have activated if the REF SPEEDS switch was not set to INCR). 1.29

Colgan Air Sterile Cockpit Policy

The Colgan Air Flight Operations Policies and Procedures Manual (FOPPM) addresses standardization, cockpit decorum, sterile cockpit concept, and airplane control on pages 5-23 to 5-25. See Operations Group Chairman Factual Report, p. 28. It states, “The captain is to maintain at all times a businesslike environment in the cockpit that is conducive to the safe and proper conduct of the flight.” Id. The FOPPM refers to 14 CFR 121.542, which states: Flight crewmember duties. (a) No certificate holder shall require, nor may any flight crewmember perform, any duties during a critical phase of flight except those duties required for the safe operation of the aircraft. Duties such as company required calls made for such nonsafety related purposes as ordering galley supplies and confirming passenger connections, announcements made to passengers promoting the air carrier or pointing out sights of interest, and filling out company payroll and related records are not required for the safe operation of the aircraft. (b) No flight crewmember may engage in, nor may any pilot in command permit, any activity during a critical phase of flight which could distract any flight crewmember from the performance of his or her duties or which could interfere in any way with the proper conduct of those duties. Activities such as eating meals, engaging in nonessential conversations within the cockpit and nonessential communications between the cabin and cockpit crews, and reading publications not related to the proper conduct of the flight are not required for the safe operation of the aircraft. (c) For the purpose of this section, critical phases of flight includes all ground operations involving taxi, takeoff and landing, and all other flight operations conducted below 10,000 feet, except cruise flight.”

37

Note: Taxi is defined as “movement of an airplane under its own power on the surface of an airport.” See Operations Group Chairman Factual Report, pp. 28-29. 1.30

Colgan Air Aviation Safety Action Program (ASAP)

Colgan Air has an ASAP program which involves its pilots, flight attendants, dispatchers, and mechanics. See Human Performance Group Chairman Factual Report, p. 20. The program was started in 2005. Id. The ASAP program coordinator manages the inflow of ASAP reports, makes sure that the ERC findings are disseminated to the appropriate personnel for corrective action, and evaluates the ASAP data to identify trends. Id. The Manager of Flight Standards, a representative from the FAA, and a pilot representative comprise the ERC. Id. Colgan Air uses an FAA contractor for the ASAP forms and to manage the database. Id. The forms used by pilots and other groups covered under ASAP are available electronically on the company’s website. Id. About 12-15 ASAP reports are submitted each month. See NTSB Public Hearing Transcript, p. 395. The database is analyzed every quarter and they look for trends and repeating events or airport locations. See Human Performance Group Chairman Factual Report, p. 20. The ASAP coordinator complies a quarterly report and also an annual report. Id. The quarterly reports are disseminated to management channels in flight standards and flight operations, and are posted on the company website for all employees to access. Id. Most ASAP data involve altitude deviations, route deviations, and slight deviations from clearances. See Human Performance Group Chairman Factual Report, p. 21. Also, ASAP data revealed some issues with overspeed during the New York ATC environment. See NTSB Public Hearing Transcript, pp. 459-60. There have been no ASAP reports regarding sterile cockpit violations in the 12 months preceding the accident, and there were no pre-accident ASAP reports on stall warning activations, severe icing, or temporary loss of control of the aircraft. See Human Performance Group Chairman Factual Report, p. 21. 1.31

Colgan Air Fatigue Policy

Colgan Air has a fatigue policy covering its pilots and flight attendants. See Human Performance Group Chairman Factual Report, p. 30. The fatigue policy and steps for calling in fatigued are covered during basic indoctrination ground school. Id. Colgan Air’s fatigue policy states: Colgan Air recognizes that there may be occasions and/or circumstances where a Crewmember’s ability to accept or complete an assignment is altered by fatigue. While our concerns are oriented to serve safe operations, we need to review all of the known factors which have led to a call of Crewmember fatigue and any resultant operations impact. This information

38

will facilitate the development of fatigue history and identity factors which led to fatigue. We can then evaluate fatigue and its relationship to operational considerations which may improve our planning and prevent recurrence. Id. When a Crewmember is unable to complete an assignment or reassignments because of fatigue, he/she must accomplish the following: Immediately notify SOC and the Operations Duty Officer. Complete the Crewmember Fatigue Form located on the following page. Within 24 hours of being released from duty because of declared fatigue, FAX or deliver the completed form to the Chief Pilot or Duty Officer at [fax number redacted]. Id. The Crewmember Fatigue Report contains entry fields for the following: name, employee number, date, local time, check-in time, local time fatigue reported to system control, block flown, time on duty today in hours and minutes, aircraft type, pairing number, bid line, reserve line yes or no, released from duty by, at what time. See Human Performance Group Chairman Factual Report, p. 30. It also contains a grid to be filled out by the crewmember listing the flight number, origination, destination, en route time, and weather, for all legs known. Id. at 30-31. Crews are asked to note legs added to the original pairing schedule. Id. at 31. The reverse side of the form contains a grid to be filled out by the crewmember listing the flight and duty time for the six days preceding the fatigue call. Id. It contains date, pairing number, bid line number, reserve yes or no, flight time, duty time, and hours rest before duty. Id. At the time of the accident, if a pilot reported fatigued, the chief pilot would be notified by crew scheduling, and would ensure that the pilot submitted the required report. See Human Performance Group Chairman Factual Report, p. 31. Fatigue reports were submitted to crew scheduling and then forwarded to the chief pilot. Id. If it was a one time event, the chief pilot would file the report and not take further action. Id. If there were repetitive instances of calling in fatigued, the chief pilot would call to ascertain if there was a problem that needed to be addressed. Id. From May 2008 (when the chief pilot assumed that title/position) to the time of the accident, about a dozen pilots called in fatigued. Id. The 3rd Quarter Safety Department Review included an item titled “Crew Rest Challenges” under the heading “issues and challenges.” See Human Performance Group Chairman Factual Report, p. 22. It contained the following bullets: complaints to FAA, increased crew declarations of fatigue, implemented restrictions on long duty days, and pairings audited since October 1 to ensure scheduling restrictions in place with no findings. Id. The listed restrictions were: no pairings built in excess of 13 ½ hour duty day, no duty

39

days greater than 14 hours without duty officer approval, and no duty days in excess of 15 ½ hours without crew agreement. Id. Colgan Air policy prohibits crew members from using a crew room to sleep overnight. A read-and-sign memo from the Newark chief pilot communicated this policy with respect to the Newark crew room and advised pilots they are responsible for their own overnight accommodations. See Read and Sign Memo 08-13 dated May 24, 2008, which reads in pertinent part as follows: SLEEPING IN OPERATIONS If a Crew Member is based in EWR then you are responsible for your own overnight accommodations. Sleeping in Operations or any crew room in EWR is strictly prohibited and will have severe disciplinary consequences, up to and including termination. See Human Performance Group Chairman Factual Report, pp. 27-28. Company records indicate that both the Captain and the First Officer acknowledged receipt of this read-and-sign memo. Id. at 27. In interviews with Colgan Air pilots, the pilots were aware of the policy against overnighting in the crew room, and none of the pilots had observed the crew room being used for overnight rest prior to going on duty. Id. at 28. 1.32

Colgan Air Pilot Scheduling

Flight time limits must not exceed 30 hours in any 7 consecutive days, 100 hours in any calendar month, or 1000 hours in any calendar year. See Human Performance Group Chairman Factual Report, p. 31. Pilots are also not scheduled for more than 8 hours flight time between required rest periods. Id. at 31-32. Pilots are required to report for duty a minimum of 60 minutes before scheduled departure time. Id. at 32. Off-duty time begins 15 minutes after block-in unless the captain notifies the company. Id. Colgan Air uses CrewTrac, a computerized system, to ensure compliance with duty and rest time regulations. See NTSB Public Hearing Transcript, p. 293. Colgan Air’s schedules are printed and published for a maximum 12 hour duty day (which would include downtime from flying). See Statement of Dan Morgan, Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009. Actual revenue operation can in some circumstances extend beyond 12 hours. Under no circumstances does the pilot’s duty day exceed 16 hours. The Manager of SOC is alerted if a crew member is going to exceed a 13 ½ hour duty day.

40

1.33

Colgan Air Commuting Policy

The 2006 Flight Crewmember Policy Handbook, page 1-12, contained the following information regarding commuting: While commuting by Flight Crewmembers is understood and accepted by the Company, in no way will commuting be deemed a mitigating factor in the Flight Crewmember’s scheduling, punctuality and demeanor. All Flight Crewmembers will be fully accountable for their timely arrival and appearance at their base. Any and all expenses incurred because of commuting will be borne by the Flight Crewmember. Flight Crewmembers should not attempt to commute to their base on the same day they are scheduled to work. See Human Performance Group Chairman Factual Report, p. 27. The Employee Handbook (Page 2-4 dated February 2005) contained the following statement about commuting in the general regulations section: Colgan Air understands it may be necessary for employees to commute, however, in no way will commuting be deemed a mitigating factor in an employee’s schedule, punctuality or demeanor. All employees will be fully accountable for on time appearance at their base (for their shift). Any and all expenses incurred because of commuting will be the responsibility of the employee. See Human Performance Group Chairman Factual Report, p. 27. Colgan policy allows pilots relief from disciplinary action if, due to unforeseen flight schedule disruption, the pilot is unable to report to duty. See Human Performance Group Chairman Factual Report, p. 26. The pilot must show they were listed on reservations for two flights to qualify. Id. at 26. The procedure can be used twice in a 12 month time period. Id. Although Colgan Air has 420 pilots, the policy is only used about six times per month, which suggests Colgan pilots are properly managing their commuting arrangements. See NTSB Public Hearing Transcript, p. 355. 1.34

Colgan Air Sick Leave Policy

Colgan pilots earn ½ day sick leave after 90 days employment and then earn ½ day sick leave a month with a maximum carryover of 30 days. See Human Performance Group Chairman Factual Report, p. 32. The Flight Crewmember Policy Handbook cites that pilots earn 1.875 hours a month of paid sick leave with a maximum carryover of 112.5 hours. Id.

41

Sick calls are made to crew scheduling. A doctor’s note is not required. Id. There is no follow-up for a pilot calling in sick unless it occurs on a repetitive basis. Id. Colgan pilots are always paid 75 hours guaranteed, regardless of events such as sick calls (unless the pilot has no accrued sick leave). See NTSB Hearing Transcript, p. 272. 1.35

Colgan Air CRM Training20

Crew Resource Management (CRM) is presented during initial new hire indoctrination as an eight hour class and during recurrent training as a two hour class. The course addresses the relationships between crew members and the use of outside resources, like system operations, gate and ramp personnel, and maintenance. Having clear communications skills is emphasized. Sterile cockpit procedures are also discussed during this training. The course uses accident case studies to raise discussion points on the benefits of good human factors and crew interaction. Issues with automation usage are also discussed to emphasize the importance of awareness of automation systems and their modes of operation. There are team-building exercises during indoctrination. The average class size is about 12 students. Dispatchers attend classes with pilots. Prior to flight attendant training being moved to Albany, there would be occasional joint pilot-flight attendant CRM classes held in Manassas. CRM is also evaluated and practiced during systems integration and simulator training and checking activities. CRM and effective communications are also stressed in company guidance outlining crew actions before departure, such as the captain conducting a crew briefing to set the tone for a positive working environment; and in the duties and responsibilities outlined for the captain to “actively promote and utilize CRM while on duty.” A 45-slide PowerPoint presentation was used in this training. The slides addressed the subject areas and modules outlined in the Crewmember and Dispatcher Training Program (CMDTP) manual and included command, leadership, leadership styles, expectations and standardization; team management, communication, situational awareness, decision making, and automation interface issues. One slide was titled “automation awareness” and contained the following bullets: situation awareness, automation management, mode awareness, energy state awareness, terrain awareness, and systems awareness. An 11-slide PowerPoint presentation on situational awareness was also used during the CRM training. One slide was titled “clues to loss of situational awareness” and contained the following bullets under the heading operational clues: failure to meet targets, undocumented procedure, departure from SOP, violating minimums or limitations, not flying airplane, and not looking outside. Another slide addressing this issue contained the following bullets: communications, ambiguity, unresolved discrepancies, preoccupation or distraction, confusion or empty feeling. The slide deck ended with the introduction of the error chain and contained a slide titled “how to break the chain” which

20

This material may be found beginning on page 13 of the Human Performance Group Chairman Factual Report.

42

included the following bullets: maintain situational awareness, checklist discipline, and standard operating procedures. Regarding the development of monitoring skills, a ground school instructor said that the CRM class discusses situational awareness and the importance for the pilot monitoring to be observant, assertive, and communicative. The CRM class provides pilots with sufficient information to develop monitoring skills. The class includes instruction on monitoring the instruments and gauges and is something that is briefed, discussed, and performed every day. Colgan Air’s Flight Operations Policies and Procedures Manual (FOPPM) contains a section outlining the use of checklists. Included in the section is the statement: While using checklists in all aircraft, it is imperative that all pilots avoid the temptation to become so engrossed in cockpit duties that outside vigilance is reduced. Also, preliminary landing and final landing checklists shall whenever possible, be completed in sufficient time so that attention to them does not distract pilots from other critical tasks. 1.36

Colgan Air Safety Culture

Colgan Air has a robust safety culture. The company has a top down commitment to safety featuring open lines of communication and constant interaction both within the Company and between the Company and the FAA. See Human Performance Group Chairman Factual Report, p. 24. This has resulted in an excellent historical safety record. The company has operated since 1991 and has flown over 10 million passengers. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 2. Prior to this tragic accident, Colgan Air never had a single passenger fatality. Id. Colgan Air has daily conference calls wherein safety is the first item on the agenda. Colgan Air also has monthly and quarterly newsletters, and quarterly Safety Review Board and Safety Council meetings to further safe operations. See Human Performance Group Chairman Factual Report, p. 17. A representative of the Safety Department attends new hire indoctrination training to provide an overview of the company’s safety programs. See Human Performance Group Chairman Factual Report, p. 17. All new-hire employees receive this orientation and are presented information about the company’s safety programs, as well as how to submit safety reports, contact safety personnel, access safety information on the company website, and find safety information in the company manuals. Id. The company holds regular meetings of senior management personnel to discuss safety. See Human Performance Group Chairman Factual Report, p. 17. These include quarterly Safety Council meetings chaired by the president/CEO of the company and attended by senior (vice president and director levels) personnel representing all company departments. Id. The purpose of the Safety Council meetings are to “[increase] safety

43

awareness, to raise unresolved safety issues, facilitate group discussion and develop positive outcomes.” Id. In addition to these meetings, the company also holds Safety Review Board meetings involving its middle management. Id. The company uses these vehicles, in part, to identify emerging safety trends. See NTSB Public Hearing Transcript, p. 404. Safety is also the top item on the agenda at daily operational meetings which include Safety Department personnel. See NTSB Public Hearing Transcript, p. 404. Safety is also the first item discussed at weekly senior staff meetings. Id. Colgan Air’s number one guiding principle outlined in its manuals is “never compromise safety.” The Employee Manual under a section titled “operational goals” states, “Safety – Our primary goal is to provide 100% safe transportation for our customers. Safety is the first priority of Colgan Air. No other value or goal has priority over safety.” The general policies and procedures section of the FOPPM begins with the statement, “In all aspects of Colgan Air operation, safety shall be given primary consideration. Each and every employee is responsible for ensuring safety in his own daily operations and shall promote safety among his fellow employees.” The company has a Safety Program Manual outlining its safety policies and procedures, and its safety reporting programs. The manual outlines the company’s nonreprisal policy which states that disciplinary actions will not be taken against anyone immediately disclosing an occurrence involving safety. See NTSB Public Hearing Transcript, p. 458. It also outlines the safety roles and responsibilities of personnel throughout the organization including the president, vice presidents, directors, managers and supervisors, and employees. Safety information is communicated to Colgan Air’s pilots via CrewTrac messages, read-and-sign memos, monthly newsletters, and safety bulletins. See Human Performance Group Chairman Factual Report, p. 18. In the CrewTrac computer system, in an addition to company emails and messages which must be reviewed and acknowledged before the pilot can check his/her schedule. Id. There is also a section maintained by the Safety Department on the crew website, along with a section maintained by the Flight Standards Department that has safety-related information. Id. Colgan Air personnel can report safety issues on a 24-hour anonymous hotline. See Human Performance Group Chairman Factual Report, p. 18. This, however, is unnecessary and seldom used, since safety issues are usually raised directly. Id. Colgan Air also has multiple forms employees can use to report safety issues, such as a Feedback Reporting Form, an aircraft/equipment/facilities damage report, a cabin safety report, a disruptive passenger report, an aviation safety action program report, a personal injury form, a ground safety and hazard reporting program form, an undeclared dangerous goods discrepancy report, and an irregularity event report form. Id. at 19. In August 2008, Colgan Air safety personnel, including the Vice President of Safety and Regulatory Compliance and the company President, traveled to all bases to conduct a

44

briefing to pilots and operational personnel on safety and the company’s safety programs (including ASAP and LOSA). See Human Performance Group Chairman Factual Report, p. 18. A 31 slide PowerPoint presentation titled “Safety an Olympic Event” was presented in what managers described as a safety road show lasting 90-120 minutes. Id. The slides discussed expectations for safety including operating the safest airline in the industry and included one in which employees express ideas openly in a participative environment. Id. They addressed safety culture and discussed industry accidents involving organizational and management factors. Id. The slides quoted an NTSB accident report stating, “the airline’s management did not instill an adequate safety orientation in its maintenance personnel by emphasizing the importance of adhering to procedures. … Senior management created a work environment in which a string of failures became probable. Accordingly, their role must be considered causal in this accident.” Id. The slides discussed methods to prevent an accident of that type, including “ensure a culture of safety exists in your company.” Id. Ways to change culture by providing top down guidance that is put into practice and reinforced were also discussed. Id. Additional accidents involving chains of errors were reviewed and the summary slides at the end stated that what causes accidents is, “rarely a single event, often a failure of a common procedure, and inattention to routine tasks and complacency.” Id. Colgan Air operations also receive periodic safety audits, including Department of Defense audits, IOSA audits, and internal evaluation program audits. Colgan Air’s safety programs go above and beyond what is required by the FAA. For example, Colgan has an ASAP program, and conducts periodic Line Oriented Safety Audits (LOSA). Colgan Air’s commitment to safety led to the adoption of additional programs such as FOQA. Starting a FOQA program is expensive and time-consuming (implementing FOQA takes about a year). Nonetheless, prior to the accident, Colgan Air made a commitment to institute FOQA as part of its ongoing commitment to safety. After the last non-QAR (quick access recorder) equipped aircraft was delivered to Colgan Air on August 1, 2008, Colgan Air authorized the funding to purchase and install QAR equipment in its fleet. Colgan Air developed an implementation and operations (I&O) plan which tracked AC-120-82 and, in October 2008, was approved by the FAA’s Principal Operations Inspector for Colgan Air, as well as FAA AFS-230 (Voluntary Safety Programs Branch). As part of this process, approved airlines such as Colgan Air agree to brief, on a quarterly basis, the POI on safety trends and corrective actions. In 2008, Colgan Air also signed a contract with a vendor for data analysis and began actively negotiating for QAR purchases. At present, Colgan Air is one of only 24 of the 101 Part 121 and/or Part 121/135 operators that have FOQA programs in place. See Interview Notes of Thomas Longridge, Operations Group Chairman Factual Report, Addendum 3.

45

2.

ANALYSIS 2.1

The Flight Crew Did Not Respond to the Unexpected Stall Warning in Accordance with Colgan Air’s Training and Prescribed Procedures

In response to the unexpected stall warning, the flight crew did not follow Colgan Air’s training and prescribed procedures. Their inappropriate flight control inputs in response to the unexpected stall warning contributed to the accident. The crew had set a Speed Bug to the standard Vref speed of 118 kts despite the REF SPEEDS switch being set to INCR, so the stick shaker activation was not expected by the crew at 130 kts. Rather than release back pressure and add power, as he was trained to do, the captain pulled back on the yoke. Less than two seconds following stick shaker activation, the control column went from 2 to 6 degrees. The nose pitched up, the angle of attack increased rapidly, and the airspeed decreased. Contrary to his Colgan Air training, the captain did not add full power after the onset of the stick shaker. Instead, the captain waited three seconds after the onset of the stick shaker before increasing power to only 75% torque. As a result, in the seven seconds following onset of the stick shaker, airspeed rapidly decreased to 100 knots, and the stick pusher activated. The stick pusher became fully active at 2216:36, when the aircraft pitch attitude was approximately 27 degrees, the angle of attack was approximately 36 degrees, and the airspeed was approximately 100 knots. The captain again pulled back on the yoke. In the three seconds following full stick pusher activation, the control column’s pitch increased over 9 degrees. The nose of the aircraft eventually dropped below the horizon, and the aircraft ended up in a severe nose low attitude from which a recovery was not possible. Runs performed by the NTSB in the FlightSafety Q400 simulator in St. Louis demonstrate that if the accident flightcrew had performed according to Colgan Air training on approach to stalls, a stall would have been avoided. See Human Performance Group Chairman Factual Report, Addendum 1, Attachment 3. Applying the proper control inputs in response to the stick shaker activation prevented the stick push from activating during the runs. Id. Further, the recoveries were uneventful and did not result in any unusual attitude situations. Id. 2.2

The Lack of Warnings Regarding the Effect of Setting a Non-Ice Vref Speed with the REF SPEEDS Switch Set to INCR Led to the Stall Warning

Flight 3407 was in icing conditions from shortly after takeoff until the accident that occurred on short final. The crew correctly activated the aircraft de-icing systems shortly after takeoff and also correctly positioned the REF SPEEDS switch to INCR at that time. However, the flight crew errantly set the #1 Speed Bug to the normal Vref speed with the REF SPEEDS switch set to INCR during the descent into BUF, which allowed for stick shaker activation at an airspeed greater than the Vref speed set in the #1 Speed Bug. The

46

fact the crew set a normal Vref speed meant they would not have expected stick shaker activation unless the airspeed had fallen below the normal Vref speed. Accordingly, the activation of the stick shaker most likely surprised the crew, and they may not have properly analyzed the stick shaker activation (i.e. the speed was too high for stall warning). Because there were no warnings or checklist items to remind the crew of the necessity to increase the approach and reference speeds (Vapp and Vref) by 20 knots, they were unprepared for the activation of the stick shaker. The “Take-off in or into Icing Conditions” section of the Q400 AFM and AOM contains a warning that states as follows: CAUTION If airspeed is not increased before REF SPEEDS switch is set to INCR, stall warning may occur. See AFM § 4.7.2.1 (Exhibit 2KK); AOM § 2.15.7.19 (Exhibit D, attached hereto). However, this warning is not set forth under the Landing in Icing Conditions section of the Q400 AFM and AOM, and thus was not applicable to the regime of flight in which the stall warning occurred. See AFM § 4.7.2.4 (Exhibit 2KK); AOM § 3.4.6.5 (Exhibit 2SS). In fact, those sections of the AFM and AOM do not even reference the REF SPEEDS switch. Further, neither the AFM nor the AOM contains a warning that if a normal Vref speed is bugged with the REF SPEEDS switch selected to INCR, a stall warning may occur at speeds greater than the non-ice Vref speed. Colgan Air has since added warnings regarding the operation of the REF SPEEDS switch to its checklists. See Section 1.27, above. Also, the act of setting a normal Vref speed with the REF SPEEDS switch set to INCR does not trigger any visual or aural warning in the Q400. Thus, the aircraft does not provide the flight crew with feedback indicating that this combination could potentially trigger a stall warning at speeds well above the bugged Vref speed. If the flight crew had received the warnings referenced above, they likely would have bugged the icing Vref speed (in this instance 138 kts. [Vref + 20 kts]) and would have added power to the aircraft sooner to stabilize the airspeed at a speed greater than the stick shaker activation speed. Further, the crew would have been more aware of the possibility of stick shaker activation. Accordingly, lack of such warnings contributed to the accident. 2.3

The Lack of a Warning or Caution Light Alerting the Flight Crew to the REF SPEEDS Led to the Stall Warning

The INCR REF SPEED display that appears on the Engine Display of the Electronic Instrument System when the REF SPEEDS switch is set to INCR is white in color, rather than amber or red. Also, the switch has no impact on any engine or engine-related

47

functions. Further, there is no corresponding display on the Warning and Caution Lights Panel. FAR 25.1322 requires that Caution Lights – lights indicating the possible need for future corrective action – be amber in color. The REF SPEEDS switch appears to be unique to the Q400 and the consequences of not following the proper procedures when the switch is set to INCR are quite serious, especially given the significant performance penalties (i.e. requirement to add 20 knots to Vref speed and increase landing field length by 25%). Therefore, the INCR REF SPEED display light should be amber and located on the Warning and Caution Panel to ensure that the flight crew is aware of its position prior to setting the Vref bug speed. The lack of appropriate caution/warning lights increases the likelihood that a flight crew may not realize the REF SPEEDS switch is set to INCR, and thus may select a normal Vref speed that will serve to raise the stick shaker activation speed above the low speed cue. If the REF SPEED selection had triggered appropriate caution/warning lights, the flight crew would have more likely to realize the switch was set to INCR and that Vref ice speeds should be bugged. Accordingly, the lack of such caution/warning lights contributed to the accident. 2.4

The Flight Crew’s Sterile Cockpit Violations May Have Caused a Loss of Situational Awareness

The accident flight crew violated Colgan Air policy and FAA regulations by breaking sterile cockpit during the descent to BUF. These violations may have distracted the crew from their duties and may have contributed to the cause of the accident. The sterile cockpit violations are as follows: Time 2205:08

Topic of Conversation Discussion of Saab procedures

2206:45

Swearing/laughing

2208:02-2208:11

Discussion of Colgan operations – other Colgan aircraft at gate

2209:26 – 2209:35

Discussion of ears popping

2210:57 – 2213:14

Discussion of flying in NE after flying in Phoenix – experience in icing when interviewed and number of hours Renslow flew before getting hired and experiences in icing

2213:58

Further comment by CA relative to above conversation

48

These violations, however, cannot be attributed to Colgan’s policies, checking, or training. Colgan Air has a clear policy against violating sterile cockpit which is part of Colgan’s training and is enforced by Colgan Air check airmen. Colgan’s policy on sterile cockpit did not contribute to the accident. Colgan’s sterile cockpit procedures are covered in ground school during general indoctrination training for pilots on the first day during review of FOPPM subjects. See Human Performance Group Chairman Factual Report, p. 28. During this review, applicable manual policies and pages are referenced. Id. Pilots are also trained on critical phases of flight in basic indoctrination. Id. at 29. Sterile cockpit procedures are expected to be followed during checkrides and they also expect to see it briefed during a captain’s preflight briefing. Id. Read-and-sign memo 08-04, dated March 6, 2008, was issued by the Director of Flight Standards on “Airspeed Awareness for Q400 Operations.” The memo stated: The Flight Operations Policy and Procedures Manual has an Altitude Callouts Table on page 5-39. The guidance is that the PF crewmember will make the 10,000 feet MSL call out. This altitude has much significance such as sterile cockpit, pressurization checks and Flight Attendant notification. With the Q400 we must now be prepared to scan the instruments for proper airspeed. See Human Performance Group Chairman Factual Report, p. 29. During post-accident interviews, one first officer stated that the company adheres to sterile cockpit procedures. See Human Performance Group Chairman Factual Report, p. 29. Another first officer said the captains he flies with very seldom deviate from sterile cockpit procedures and he has spoken up when they do. Id. When these rare violations occur, he directs the captain to the business at hand, such as asking about the descent check or the approach briefing, to remind them they are in the sterile cockpit period. Id. Check airmen and captains also stated that the pilots they fly with maintain sterile cockpit. Id. Some said if pilots forget they will correct the situation by reminding them they are under sterile cockpit. Id. Others said that ignoring a statement conveys the proper message. Id. The FAA APM stated he had not seen any “red flags” suggesting Colgan had any problem with sterile cockpit before the accident. See Human Performance Group Chairman Factual Report, p. 29. Similarly, the POI stated that FAA surveillance of Colgan before the accident did not indicate there was a problem with crew adherence to sterile cockpit procedures; and based on his knowledge of the company’s own oversight findings there has been no indication of a problem either. Id. Colgan Air uses irregularity reports for reporting violations of standard operating procedures such as sterile cockpit violations. See NTSB Public Hearing Transcript, pp. 217-18. Reports are discussed between the Director of Operations and the Chief Pilot. Id.

49

The offending flight crew is removed from flight status and is reported to Flight Standards. Id. Incidents of violation are few. 2.5

The Flight Crew Did Not Adhere to Colgan Air Policy on When Checklists Should Be Accomplished

The Q400 CFM states in Section Five, “normally, the PF calls for the DESCENT checklist when descending through 18,000’ or at the top of descent if cruise altitude is below 18,000’.” The Descent Checklist reads as follows: Altimeters …………………….. Fuel Balance …………………. Pressurization ………………... Cabin PA ………………………

SET/CROSSCHECK CHECK SET COMPLETE

CR PM PM PM

See CFM, Section 5. Here, cruise altitude was 16,000 feet. Captain Renslow, however, did not call for the descent checklist until 2,980 feet. The CFM also provides that The PF calls for the APPROACH checklist with sufficient time to complete it prior to crossing the initial approach fix or transitioning to the initial approach phase during an instrument approach. Here, Captain Renslow called for the approach checklist immediately after completion of the descent checklist. The approach briefing was not done as part of the check, but instead was done during descent from 11,000 at 2204-2205. The flight crew’s deviations from checklist discipline, sterile cockpit, and other standard operating procedures are very unusual for Colgan Air pilots. Line checks conducted by Colgan Air check airmen and FAA APDs prior to the accident found very high compliance with FARs and Colgan Air procedures. Flight observations conducted on all Colgan Air pilots following the accident similarly showed high rates of compliance and only minor deviations. 2.6

FO Shaw Did Not Properly Manage Her Personal Schedule Preceding the Accident

The facts regarding FO Shaw’s overnight commute demonstrate that even though she had three days off prior to February 12th, she did not plan her personal time properly prior to reporting for duty. Rather than commuting to EWR on Feb. 11 and staying in a hotel, she chose an overnight commute.21 21

Linda Morris, FO Shaw’s mother, stated that FO Shaw planned to use hotels to obtain rest. See Human Performance Group Chairman Factual Report, Addendum 1, pp. 4-5.

50

It appears that FO Shaw did not have time for adequate sleep in the 24 hours preceding the accident. However, she also took a lengthy nap on the morning of the accident flight, and said she was “feeling great.” See Human Performance Group Chairman Factual Report, Addendum 1, p. 6. No one noticed she was fatigued or ill. Based on comments of FO Shaw on the CVR, however, it is apparent that her health deteriorated at some point prior to the flight, yet she chose to fly rather than call in sick. Despite these facts, it is unclear if fatigue or illness played a role in the accident. During the flight, FO Shaw did not state she was fatigued or had trouble staying awake. Further, review of the CVR transcript shows FO Shaw was alert and responsive to Captain Renslow, ATC, and dispatch communications during the 53 minute flight. 2.7

Captain Renslow May Not Have Properly Managed His Personal Schedule Preceding the Accident

Based on post-accident interviews and review of phone records, it appears Captain Renslow may not have managed his personal time appropriately in the days preceding the accident and, as a result, may have been fatigued. After finishing a trip on February 5, Captain Renslow had the next five days off work. See Human Performance Group Chairman Factual Report, p. 6. On February 9, 2009, he spent the day at home in the Tampa area. Id. At 1713, he departed Tampa and arrived in EWR at about 2005. Id. At 2247, he made a 6 ½ minute phone call. Id. The first officer who flew with Captain Renslow on February 10th stated that Captain Renslow spent the night of February 9th in the crew room at EWR. See Human Performance Group Chairman Factual Report, p. 6. It is against Colgan Air policy to spend the night in the crew room. A hotel or a shared apartment is a more appropriate place to receive quality rest. Further, Captain Renslow’s report time on February 10th was 0530. Id. Since Captain Renslow was apparently on a phone call shortly before 11:00 p.m. on February 9th, it appears he did not receive more than six hours of sleep. On February 10, Captain Renslow had a duty day of 7:49 (hh:mm), which included 4:36 of flight time. Id. The trip ended at BUF at 1259. Id. Captain Renslow had the rest of the day off. Id. According to the first officer who was flying with Captain Renslow that day, they spent the afternoon and evening relaxing at the hotel. Id. The first officer last saw Captain Renslow about 2100-2130 as they left the pool-community area of the hotel to go to their rooms. Id. Phone records indicate the captain made a 30 minute call at 2102. Id. On February 11, the first officer reported seeing Captain Renslow in the hotel breakfast area at approximately 0500. Id. Accordingly, even though the captain had a 16:56 rest period, he likely did not receive more than seven hours of sleep.

51

Captain Renslow’s duty day on February 11 was 9:49, which included five hours of flight time. See Human Performance Group Chairman Factual Report, p. 7. His last flight that day arrived at EWR at 1544. Id. During post-accident interviews, the first officer said Captain Renslow had told him he was going back to his apartment that evening. Id. The first officer stated the captain’s health appeared very good and the captain was well rested and alert during the flights on February 10 and 11. Id. On February 11, Captain Renslow placed or received calls between 1552 and 1637, 1823 to 1829, and 2020 to 2142. See Human Performance Group Chairman Factual Report, p. 7. Crew Trac logon records indicate that he had multiple logons from 1610 to 1640, 1759 to 1831, and 2059 to 2110. Id. The last logon occurred at 2151. Id. Accordingly, Captain Renslow likely did not fall asleep until after 2200 on February 11th. On February 12, at 0310, Captain Renslow logged onto the CrewTrac system, accessed the crew menu and acknowledged a revision to the accident trip’s schedule. See Human Performance Group Chairman Factual Report, p. 7. Captain Renslow logged on again at 0726, and used his phone several times between 1000 and 1100. Id. Between 1200-1400, he volunteered to perform some office work for the regional chief pilot. Id. At 1624, he made a call and at 1649 a call was received. Id. Other calls were made in the early evening hours before the accident flight. Id. at 7-8. While Captain Renslow may have been napping periodically in the early morning hours of February 12th and at other times during the day, his primary rest period for the night of February 11 was likely less than five hours. Despite having ample time off duty in the three days preceding the accident, it appears Captain Renslow averaged less than six hours of overnight sleep in that time period. Accordingly, it is possible he was fatigued during the accident flight. 2.8

Colgan Air Has a Careful, Robust Hiring Process

Colgan Air’s hiring process did not contribute to the accident. As set forth in Section 1.15, above, Colgan Air has a rigorous, multi-tiered evaluative process which Captain Renslow and First Officer Shaw successfully completed. Colgan Air pilots meet the same high, federally-mandated standards as pilots at major air carriers. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 4. All Colgan Air captains hold an airline transport pilot rating, the highest level of pilot certification offered by the FAA. Id. at 5. Colgan Air pilots often come from major flying schools, bridge programs, or other Part 121 carriers. At least two-thirds of the pilot applicants who are initially contacted for interviews are not offered employment. See Operations 2H, Interview of Vice President of Administration Mary Finnigan, p. 46; Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 4. Pilots who do get hired become training candidates. See NTSB Public Hearing Transcript, pp. 311-15; Statement of Dan Morgan before the

52

Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 4. As a training candidate, the pilot undergoes ground training, simulator training, and checkrides. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009; NTSB Public Hearing Transcript, pp. 311-15. While the candidate is in training, the PRIA paperwork is received and reviewed. Id. The pilot does not begin line service until he/she passes the PRIA check, criminal records check, national driver registry check, ground training, simulator check, and a checkride. Id. Captain Renslow and FO Shaw both successfully completed the hiring process and were successful training candidates. As set forth above, Captain Renslow was not truthful in his application to Colgan Air and failed to disclose two prior certificate disapprovals. Colgan Air complied with PRIA, but PRIA did not cover these disapprovals because Captain Renslow was not working as a pilot at the time. Colgan Air also followed all FAA recommendations that pertain to pilot hiring. At the time Renslow was hired, the FAA had not provided any guidance on the use of FOIA to obtain additional pilot information Prior to the accident, the FAA did not have clear procedures in place for obtaining Notices of Disapproval, and not a single air carrier was using FOIA. 2.9

Colgan Air Has a Thorough and Effective Pilot Checking Program

Captain Renslow failed two checks at Colgan Air (Recurrent Proficiency Training in SF-340 (10-17-06) and ATP for SF-340 (10-15-07). See Operational Group Chairman Factual Report, p. 5. Both times he was successfully retrained. He passed four training events and six checking events in the 16 months between his last failed check and the accident, including checks administered by FAA APDs. See NTSB Public Hearing Transcript, p. 296. First Officer Shaw had a good checking history. She did not fail any checking events at Colgan Air. See Operations 2BB. 2.10

Colgan Air Q400 Training Program is Robust

Colgan’s Q400 training program was not a factor in the accident. Colgan Air had a robust, FAA approved, Q400 training program. Colgan’s program is closely modeled on the training program used by Flight Safety International, which is the manufacturer’s designated flight training company. See Operations Group Chairman Factual Report, p. 35. Colgan’s training program was reviewed and approved by the FAA. Colgan uses the flight simulators of Flight Safety International for its Q400 trainees as well as other state of the art equipment such as a flight management system trainer and a ground flight simulator. Id.; NTSB Hearing Transcript, p. 158. Before Colgan Air pilots can operate a Q400 aircraft as a fully-qualified crew

53

member, they must complete 156 hours of Q400 flight and ground training, including: a. b. c. d. e. f. g. h. i. j. k.

normal and abnormal procedures training; stall recovery procedures training; winter operations training; sterile cockpit procedures; aircraft systems; standard operating procedures; an online Flight Management System training course; three cockpit procedures training sessions; four observation flights with experienced Q400 pilots; at least 20 hours flying with a check airman observing; and two successful check rides

Before Captain Renslow flew a single passenger as Pilot in Command in the Q400, he received: a. 156 hours of Q400 flight and ground training; b. four observation flights with experienced Q400 pilots; c. over 20 hours flying with a check airman observing him; and d. two successful check rides Colgan Air’s captains have, on average, over 4,600 hours of flight time. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, U.S. House of Representatives, June 11, 2009, p. 5. Every captain has an Airline Transport Pilot rating, which is the highest level of pilot certification available. Id. All pilots are “type-rated” on the specific aircraft they fly, and all ratings are issued by the FAA. Id. Captain Renslow had 3,379 total hours of flight experience and was Airline Transport Pilot rated and held a “type certificate” in the Q400. See Operations Group Chairman Factual Report, pp. 4, 6. First Officer Shaw had 2,244 total hours of flight experience. She had 774 hours flying the Q400 aircraft, qualifying her fully in accordance with all applicable Federal Aviation Regulations. See Operations Group Chairman Factual Report, p. 7. Colgan Air’s Captain Upgrade curriculum includes CRM, situational awareness, and leadership training. See Human Performance Group Chairman Factual Report, pp. 15-16. FO Shaw was trained on the Q400 at Flight Safety International in Toronto, Canada. A first officer who trained with FO Shaw in Toronto said FO Shaw had good knowledge of the airplane. See Human Performance Group Chairman Factual Report, p. 8. The check airman who conducted the IOE in March 2008 said FO Shaw did well and he did not recall airplane handling issues or her struggling to keep up with the airplane. Id. at 8-9. He described her as a good pilot, pretty sharp, assertive and thorough. Id. at 9.

54

FO Shaw had a positive reputation among captains she had flown with for being a pilot who did a good job and performed as expected. See Human Performance Group Chairman Factual Report, p. 9. A captain who had flown with her multiple times said she was average to above average for her level of experience. Id. He described her as proficient, and as a monitoring pilot she was always ahead of the airplane, not complacent, and she cross checked what she did. Id. In post-accident interviews, several captains remarked that based on FO Shaw’s abilities she would have upgraded to captain in time. See Human Performance Group Chairman Factual Report, p. 9. She was characterized as assertive and not overly talkative or chatty in the cockpit. Id. None of the captains who were interviewed reported any difficulties with her adhering to sterile cockpit procedures, nor did they observe her making any configuration changes to the airplane unprompted while they were the flying pilot. Id. The ground school instructor for FO Shaw’s recurrent training in January 2009 described her as enthusiastic and attentive, and thought she had good knowledge of the airplane based on questions she answered. Another first officer in that class described the accident first officer as being more experienced in technical knowledge than the average first officer. See Human Performance Group Chairman Factual Report, p. 9. 2.11

Colgan Air Provides Appropriate Remedial Training

FAA Notice 8900.71 prescribes remedial training for Part 121 operators. The Notice directs the FAAs Principal Operation Inspectors to determine if operators under their surveillance had voluntarily accomplished the recommended action listed in FAA SAFO 06015, the purpose of which was to promote voluntary implementation of the remedial training programs for pilots with persistent performance deficiencies. See Human Performance 14I. Specifically, the SAFO recommended that the directors of safety, whose carriers did not have such programs, to recommend a process to top managers that would identify pilots who have persistent performance deficiencies and who have experienced multiple failures in training and checking. Id. The SAFO suggested that the three objectives be accomplished by the operator: 1) review the entire performance history of any pilot in question; 2) provide additional remedial training, as necessary; and 3) provide additional oversight to ensure that performance deficiencies are effectively addressed and corrected. Id. Colgan Air provides remedial training for crewmembers with unsatisfactory checkrides. See Operations Group Chairman Factual Report, p. 40. If a pilot has performed unsatisfactorily, the Director of Flight Standards notifies Training.and establishes a plan to provide additional training. Id. If the pilot does not pass the recheck, he/she may be terminated. Id. Also, the Chief Pilot reviews pilots with prior training deficiencies to determine if they are suitable for upgrade and transition. See Operations Group Chairman Factual Report, Addendum 3, Attachment 6.The Chief Pilot also speaks with the candidate to satisfy himself that the candidate is suitable for upgrade/transition

55

and to inform the candidate of any further checking that will be done prior to upgrade or transition. Id. In addition, the FAA is notified by Colgan Air whenever a pilot fails a checkride. The failure is passed to the FAA air crew program manager for the particular fleet type (Saab 340 or Q400), who will sit in on the re-checks if they believe it is necessary. Following the accident, Colgan Air reviewed the checking history of all its 420 pilots and found that only six had more than one check ride failure at Colgan Air. See NTSB Public Hearing Transcript, p. 204. 2.12

Colgan Air Provided the Flight Crew with Proper and Thorough Stall Training

Captain Renslow and First Officer Shaw received the required stall training that was part of Colgan’s FAA approved training program for the Q400. Colgan meets all FAA requirements for stall training, and Colgan’s stall recovery procedures follow the FAA’s Practical Test Standards. Captain Renslow and First Officer Shaw also received substantial stall warning training and checking prior to operating Q400s at Colgan Air. For example, a pilot must pass a certification check ride in which stalls are a required maneuver to obtain each grade of license (i.e., private pilot, instrument rating, commercial rating, and multi-engine rating). Therefore, prior to entering a Part 121 training program, every single pilot must have shown proficiency in recovering from a stall during four certification events. Captain Renslow would have demonstrated the proper recovery procedure of releasing back pressure on the yoke, instead of pulling back, at the initiation of the stick shaker, during six certification events at a Part 121 airline (in addition to his 12 training sessions). Captain Renslow performed at least 35 stall recovery procedures during his career at Colgan Air. He was also trained in stall recovery at Gulfstream International Airlines. The stall training and checking Captain Renslow received throughout his aviation career is important because stall recovery procedures are the same for every aircraft – decrease the angle of attack, add power, and correct the roll. Colgan Air’s FAA approved Q400 training program did not include stick pusher training. Stick pusher simulator training, however, is not a standard practice in the airline industry. If a pilot responds appropriately to the stick shaker, he/she will not experience the stick pusher. Further, Bombardier does not publish a procedure for recovery from a deep stall, only stall warning recovery. Like Colgan Air, other Q400 operators, such as Porter Airlines, taught recovery from stall warning rather than deep stall.22 See Operations Group Chairman Factual Report, Addendum 3, Attachment 4.

22

Colgan also had simulator training including upset recovery training (nose low, nose high, high bank angles, etc.).

56

In addition to his training in the Q400, Captain Renslow received stall training in the Saab 340, which included stick pusher training. He would have felt operation of the Saab stick pusher on almost 100 occasions. See Spreadsheet of Captain Renslow’s Originating Flights in Saab 340 and Q400. As set forth in Section 1.24, above, the stall warning system for the Saab 340 and Q400 include the stick shaker and the stick pusher, and the operation of each is very similar. The force exerted by the stick pusher in the Saab 340 and Q400 is also quite similar. The fact that the Saab 340’s stall warning system (including the stick pusher) is so similar to that of the Q400 means that operation of the Q400 stick pusher would not have been a novel sensation to Captain Renslow. The fact that Colgan Air’s stall warning training was not a factor in the accident is apparent from test runs performed by the NTSB. These runs demonstrate that if a pilot responds to a stall in accordance with Colgan Air’s stall training, a Q400 stall will be quickly corrected. See Human Performance Group Chairman Factual Report, Addendum 1, Attachment 3. The stick pusher did not activate during any of the NTSB’s test runs. Id. The recoveries were uneventful and did not result in unusual attitudes. Id. During the accident flight, Captain Renslow did not follow Colgan Air’s training on approaches to stalls. In response to the stick shaker, Captain Renslow applied only 75% power, and pulled back on the yoke, rather than reducing the angle of attack and adding full power. It is unclear why Captain Renslow did not follow his training. Colgan Air’s stall training is not a factor in this accident. 2.13

Colgan Air Provides Appropriate Winter Operations Training

According to Bombardier’s testimony at the public hearing, the Q400 cannot tail stall. Colgan Air does not train its pilots on tail stall recognition or recovery, nor does Colgan Air have any written guidance or prescribed procedures for tail stall recognition or recovery. It is not possible to train for tail stalls in the simulator, and Colgan Air pilots are not exposed to any tactile or hands-on experience with tail stalls. Colgan Air, however, does provide its pilots with winter operations training. Part of this training includes a NASA video that is used for general education (Icing for Regional & Corporate Pilots). This video is also shown in initial and recurrent ground school. Captain Renslow and FO Shaw both saw the NASA video as part of their training. See Operations Group Chairman Factual Report, p. 38. Their most recent training events that included the NASA video were as follows: o Captain Renslow completed transition winter operations anti-ice/de-ice training on October 31, 2008, during transition training for the Q400. o FO Shaw completed recurrent winter operations anti-ice/de-ice training on January 15, 2009.

57

The NASA video discusses the latest information from NASA and the FAA about icing. The purpose of the video, as stated in its introduction, is: o to review fundamentals of aircraft icing o to enhance the pilot’s ability to assess hazardous icing conditions o to enhance the pilot’s understanding of icing effects on stability and control of the aircraft o to present strategies that pilots can use to exit a hazardous icing encounter o to discuss super cooled large droplets (SLD) One subject the video also discusses is the possibility of horizontal stabilizer icing and associated tail stall. It explains differences between conventional wing stall and tail stall. The video states, in part, that in the event of a wing stall, recovery requires reduction of the angle of attack, which is accomplished by lowering the nose, adding power, and lowering flaps one notch. It also states that pilots should be alert to the warning signs of a tail stall. These include:     

Lightening of the controls Pitch excursions Difficulty in trimming pitch Buffeting of the controls Sudden nose down pitch

The video states that to recover from a tail stall, a pilot must:   

Pull back on the yoke Reduce flaps Reduce power (some aircraft)

The video states that the pilot must first properly diagnose the problem. It states that airspeed awareness is absolutely critical and, although the differences between wing stall and tail stall can be subtle, the recovery techniques are quite different. At the public hearing, a witness from Bombardier testified that the Q400 cannot tail stall. See Testimony of Allan Paige, NTSB Hearing Transcript, p. 102.23 Another Bombardier witness testified that he conducted the certification test flying for the Q400 and determined the aircraft did not have any tailplane stall tendencies. See Testimony of Wally Warner, NTSB Hearing Transcript, pp. 144-48. However, prior to the time of the accident, 23

Mr. Paige testified that he was Bombardier’s lead engineer in the flight sciences department and was in charge of assessing and evaluating the handling characteristics of the Q400. See Testimony of Allan Paige, NTSB Hearing Transcript, p. 42.

58

there was no pronouncement from Bombardier stating definitively that the Q400 cannot tail stall. See Testimony of Allan Paige, NTSB Hearing Transcript, p. 103. In fact, tail stall guidance is still included in Bombardier’s Q400 AOM. See AOM 3.4.6.5, Operations 2SS, which states in pertinent part as follows: If landing flap selection is accompanied by stick force lightning or stick force irregularities, immediately retract flap to lesser setting. Cycle wing/tail de-icer boots several times and if possible, land using a smaller flap setting. (This condition, which is the precursor to tail stall, will not occur if AOM icing procedures are followed). The Bombardier witness did not know why this language was included in the AOM. See Testimony of Allan Paige, NTSB Hearing Transcript, p. 102. Bombardier has not yet clarified in writing that the Q400 cannot tail stall. Horizon Airlines had a note regarding tail stall in its Q400 manual which stated that if the controls lightened when selecting landing flaps, the aircraft could be in a tail stall. Since the accident, they have removed that note. See Interview notes of Don Wiens, Addendum 3 to Operations Group Chairman Factual Report. Lynx Aviation, another Q400 operator, had a brief word-for-word excerpt from the Bombardier AOM in their manual, but did not train on tail stalls and did not show the NASA video in their ground school training. Some pilots, however, circulated the video amongst themselves. See Interview notes of Ryan Peck, Addendum 3 to Operations Group Report. Porter Airlines did not refer to tail stall in their manuals or procedures, but did mention it as a general subject in their ground school. That element has been removed from ground school since the accident. Colgan Air did not train its pilots on tail stall recognition or recovery. There are no references to tail stall or tail stall recovery techniques in the Colgan Air Crew Member and Dispatcher Training Program Manual. There is no indication the accident aircraft suffered a tail stall. Colgan Air’s training is solely geared toward wing stalls. Further, the Q400’s stall protection system is only activated by a wing stall. 2.14

FO Shaw Had Substantial Winter Operations Experience

Both Captain Renslow and FO Shaw had extensive experience flying in icing conditions.24 At the public hearing, there was discussion of remarks made by FO Shaw regarding her experience in icing conditions that were misinterpreted. On the CVR, FO Shaw referred to not having experience with icing conditions when initially hired. Her later remarks clarified that her time with Colgan Air had given her such experience, “… now I’m so much more comfortable with it all.” See CVR Factual Report – Addendum, p. 12-106. 24

The flights of FO Shaw, rather than those of Captain Renslow, are emphasized, as FO Shaw was the less experienced pilot and a misinterpretation of the CVR transcript has caused unwarranted speculation that FO Shaw was not experienced in icing conditions.

59

A review of weather for the flights operated by FO Shaw from November 1, 2008, to the time of the accident shows that FO Shaw flew in icing conditions on approximately 30 occasions either in the takeoff/climb, cruise, or approach/landing phase. See November 20, 2000 letter to Lorenda Ward. Many of these flights involved snow or freezing rain. Id. The flights are as follows: November 13, 2008 November 13, 2008 November 20, 2008 November 20, 2008 November 27, 2008 December 10, 2008 December 10, 2008 December 12, 2008 December 12, 2008 December 12, 2008 December 16, 2008 December 16, 2008 December 19, 2008 December 19, 2009 December 21, 2008 December 26, 2008 December 26, 2008 January 6, 2009 January 6, 2009 January 11, 2009 January 18, 2009 January 18, 2009 January 19, 2009 January 19, 2009

EWR to BUF EWR to DCA EWR to BUF BUF to EWR EWR to BUF EWR to BUF BUF to EWR ORF to EWR EWR to ORF ORF to EWR BUF to EWR EWR to ORF EWR to ORF ORF to EWR EWR to ORF ORF to EWR EWR to ORF ORF to EWR EWR to ORF ORF to EWR ORF to EWR EWR to ORF ORF to EWR EWR to PIT

Flight 3401 Flight 3405 Flight 3401 Flight 3404 Flight 3401 Flight 3401 Flight 3404 Flight 3223 Flight 3220 Flight 3227 Flight 3404 Flight 3224 Flight 3349 Flight 3223 Flight 3228 Flight 3227 Flight 3228 Flight 3221 Flight 3222 Flight 3223 Flight 3227 Flight 3228 Flight 3225 Flight 3236

January 19, 2009 January 19, 2009 January 20, 2009 January 20, 2009 February 3, 2009 February 4, 2009

PIT to EWR EWR to ORF EWR to PIT PIT to EWR EWR to BUF EWR to BUF

Flight 3239 Flight 3228 Flight 3236 Flight 3239 Flight 3407 Flight 3407

approach/landing approach/landing (possibly also in cruise) approach/landing (possibly also in cruise) takeoff/climb approach/landing approach/landing (possibly also in cruise) takeoff/climb and cruise takeoff/climb and possibly in cruise takeoff/climb and possibly in cruise approach/landing and possibly in cruise approach/landing takeoff/climb, enroute, and approach/landing takeoff/climb (possibly also in cruise) approach/landing takeoff/climb, descent/landing, and maybe cruise takeoff/climb and approach/landing cruise and approach/landing takeoff/climb and cruise cruise and approach/landing cruise and approach/landing takeoff/climb, cruise, and approach/landing takeoff/climb, cruise, and approach/landing cruise and approach/landing takeoff/climb and cruise (possibly approach/landing) takeoff/climb, cruise, and approach/landing takeoff/climb, cruise, and approach/landing approach/landing takeoff/climb takeoff/climb approach/landing and possibly in cruise flight

Id. FO Shaw also may have encountered icing conditions on other flights, such as the following: November 20, 2008 EWR to ORF November 21, 2008 ORF to EWR December 17, 2008 ORF to EWR

Flight 3224 Flight 3227 Flight 3227

possibly in cruise flight possibly in cruise flight possibly in cruise flight

Id. In addition to this weather data, statements of captains who flew with FO Shaw in icing conditions confirmed her experience. Id. Captain Robert Golden flew with FO Shaw 60

for the majority of November 2008. Id. Captain Golden stated that he and FO Shaw often encountered icing conditions and FO Shaw responded to such conditions appropriately and followed Colgan Air’s procedures. Id. He also recalled that FO Shaw recommended changes in altitude to potentially get out of icing conditions. Id. Captain Sam Omair flew with FO Shaw during here initial operating experience (IOE) in March 2008. Id. He recalled that they flew in icing conditions and that FO Shaw was trained for those conditions, including operation of the deicing equipment and the reference speeds switch. Id. The following IOE flights were conducted in icing conditions: March 19, 2008 March 20, 2008 March 20, 2008 March 21, 2008 March 22, 2008

BWI to EWR ALB to EWR ROC to EWR EWR to ROC EWR to PIT

Flight 3200 Flight 3221 Flight 3253 Flight 3250 Flight 3343

approach/landing, possibly in cruise flight takeoff/climb and cruise takeoff/climb and cruise possibly in cruise flight approach/landing and cruise

Id. Captain John Dowd flew with FO Shaw in December 2008. Id. He recalled flying with her in icing conditions on several occasions. Id. Captain Dowd recalls that they performed the normal deice procedures and that FO Shaw never had any problems. Id. Captain John Miller was paired with FO Shaw for January 2009. Id. He recalls flying with FO Shaw in icing conditions and that FO Shaw always followed correct procedures. Id. He recalls they landed once in icing conditions so the REF SPEEDS switch was left in the INCR position. Id. Other times they flew clear of icing conditions and turned the switch off. Id. Captain William Isbell flew with FO Shaw in early February 2009, just days before the Flight 3407 accident. Id. He recalls they encountered snow and icing on February 4, 2009, on Flight 3407, and that the weather FO Shaw did not have any problems with deicing procedures. Id. Captain Craig Margolin flew with FO Shaw in early November 2008. Id. On one of the trips, they encountered icing at cruising altitude. Id. He asked FO Shaw to look at her wing and tell him if the boots were shedding the ice. Id. She replied that the boots were working, and they both continued to check the wings periodically to make sure the wings were being cleared. Id. She never seemed nervous during this icing event. Id. This flight was either November 6, 2008, Flight 3401 from EWR to BUF, or November 6, 2008 Flight 3404 from BUF to EWR. Id. The icing event occurred closer to BUF. Id. A pilot undergoing training would not be able to sit for a practical test if he had not completed and shown proficiency in the training module for procedures relating to the ice protection system and any malfunctions associated with that system. In November 2008, Colgan Air issued a Winter Operations Quiz to all pilots on all

61

fleets. This was a quiz relating to general operations in a winter environment and had questions about the current revision of Chapter 7, Winter Operations, of the Flight Operations and Procedures Manual. Winter Operations are covered during proficiency checks in the simulator twice a year for Captains, and once a year for First Officers, and also in recurrent training (ground school) once a year. See NTSB Public Hearing Transcript, pp. 210, 249. Both Captain Renslow and FO Shaw were well trained in winter operations. Colgan Air pilots receive winter operations training to optimize their ability to operate safely in winter conditions. See NTSB Public Hearing Transcript, pp. 210-11. Both basic indoctrination and recurrent training covers winter operations generally and icing conditions specifically. Id. at 249-51. These areas are covered in ground training and in the simulator. Id. at 210, 249. 2.15

Colgan Air Pilot Scheduling Affords Ample Time for Pilot Rest

Colgan Air follows the duty and rest time regulations of the FAA. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, United States House of Representatives, June 11, 2009, p. 7. Colgan Air uses an automated crew scheduling system that tracks duty time and ensures compliance with the FAA’s duty limitations and rest requirements. Id. Crew boards are pulled twice a day to analyze schedules. See Human Performance Group Chairman Factual Report, p. 22. The Director of SOC and the Manager of SOC are alerted if a crewmember is going to exceed 13 ½ hours of duty time. Id. Colgan Air develops crew schedules to provide ample rest between duty days, and to allow periodic extended rest periods. For example, after a three or four consecutive day duty period, a pilot may have four or five days off. Colgan’s monthly schedules are determined well in advance of the beginning of each month, which affords pilots ample time to be fully rested. The schedules for Q400 pilots are built with maximum duty days of 12 hours, and with a maximum of 7 ½ hours of scheduled flight time during that duty period. See Statement of Dan Morgan before the Subcommittee on Aviation, Committee on Transportation and Infrastructure, United States House of Representatives, June 11, 2009, p. 7. For example, the average for all scheduled flight and duty times for January 2009 was 4:44 and 8:59 respectively, with an average of 13 days off each month. Id. Although sixteen-hour duty days are legal under FAA regulations, they are rarely assigned by Colgan Air. See NTSB Public Hearing Transcript, p. 330. In fact, duty days that exceed 14 hours must be reviewed by a senior operations manager prior to assignment, and Colgan Air will not assign a crew a duty day that exceeds 15 ½ hours unless the crew agrees. See Human Performance Group Chairman Factual Report, p. 22. Regarding the accident flight crew, Captain Renslow had 27 hours off between

62

the accident flight and his last flight prior to the accident flight. First Officer Shaw had four days off duty before the accident flight (her last flight prior to the accident flight was in the afternoon of February 8). 2.16 Colgan Air Fatigue Policy Promotes Safety Colgan Air has a simple, clear cut policy regarding fatigue – if you’re fatigued, you don’t fly. See FOPPM 3-124. You must remove yourself from duty. Id. Colgan Air’s policy is also non-punitive. See NTSB Hearing Transcript, p. 233. The process for calling in fatigued is straightforward and taught to all pilots in basic indoctrination ground school. See NTSB Hearing Transcript, p. 23. This policy, and Colgan Air’s emphasis on it, helps ensure pilots make the right decision not to fly fatigued. Colgan Air’s pilot schedules are generally comprised of several days working followed by several days off. It is not a five day per week, four weeks per month job. Instead, a pilot may work four days and then have four days off. In addition, crew schedules are provided at least nine days in advance, so the pilots have plenty of time to schedule themselves appropriately. For example, Captain Renslow had 22 hours off before reporting for duty on the day of the accident, and First Officer Shaw had three full days off. Colgan Air uses CrewTrac, a computerized system, to ensure compliance with duty and rest time regulations. See NTSB Hearing Transcript, p. 293. Colgan pulls crew boards twice a day to analyze schedules, and strives to keep its pilots within a 14 hour day. See Human Performance Group Chairman Factual Report, p. 22; NTSB Hearing Transcript, p. 330. Colgan Air’s Director of SOC and Manager of SOC are alerted if a crewmember is going to exceed 13 ½ hours. Id. Colgan Air has operational policies that also reduce fatigue risk. For example, Colgan schedules several pilots on “hot reserve” at each crew base. See NTSB Hearing Transcript, p. 330. Any pilots that need to be taken off the line are removed and supplemented with a reserve pilot, and can make up the flight once properly rested. See NTSB Hearing Transcript, p. 342. Thus, the pilots understand that fatigue calls will not have an adverse operational effect. Colgan Air also requires its crew members to have proper accommodations for rest at their base (such as shared commuter apartments). Accordingly, Colgan Air has a policy against use of crew rooms for overnight sleep. See Human Performance Group Chairman Factual Report, p. 27. A read-and-sign memo from the Newark chief pilot communicated this policy with respect to the Newark crew room and advised pilots they are responsible for their own overnight accommodations. Read and Sign Memo 08-13 dated May 24, 2008, reads in pertinent part as follows: SLEEPING IN OPERATIONS

63

If a Crew Member is based in EWR than you are responsible for your own overnight accommodations. Sleeping in Operations or any crew room in EWR is strictly prohibited and will have severe disciplinary consequences, up to and including termination. See Human Performance Group Chairman Factual Report, p. 28. Company records indicate that both the captain and the first officer acknowledged receipt of this read-and-sign memo. In interviews with Colgan Air pilots, the pilots were aware of the policy against overnighting in the crew room, and none of the pilots had observed the crew room being used for overnight rest prior to going on duty. Id. Pilot education also protects against fatigue. Issues of fatigue are addressed in the Airman’s Information Manual and are part of every pilot’s basic education. See NTSB Hearing Transcript, p. 234. Colgan Air reinforces the importance of not flying fatigued during ground school. See NTSB Hearing Transcript, p. 23. Also, in Captain Management Training, situational awareness and fitness for duty are discussed. See NTSB Hearing Transcript, p. 234. Colgan Air’s CRM program also provides fatigue management tools, such as factors causing fatigue, symptoms of fatigue, sleep needs and napping, and tips for getting sleep. See Human Performance Group Chairman Factual Report, p. 15. Colgan Air’s Dispatch Release serves as another reminder of the importance of being well rested. A captain must certify on the dispatch release that he or she is physically qualified for the flight. See NTSB Hearing Transcript, p. 342. A fatigued pilot is not physically qualified. In sum, Colgan Air’s fatigue policy did not contribute to the accident. 2.17

Colgan Air Commuting Policy Was Not a Factor in the Accident

Like every airline, Colgan Air has pilots who commute. It is widely acknowledged that significant numbers of airline pilots commute to work. Making appropriate commuting arrangements is part of being a professional pilot. As FAA Administrator Babbitt pointed out following a recent incident, lack of professionalism cannot be regulated. An airline cannot monitor and control unprofessional uses of personal time, whether the same involves staying up late before an early morning flight, or traveling away from base on days off for personal reasons. Colgan Air expects its pilots, and all its employees, to present fit for duty, regardless of where they reside. A commuting pilot is expected to report for duty in a timely manner. See Flight Crewmember Policy Handbook, pp. 1-5 and 1-6. Colgan Air requires its pilots to have appropriate sleeping accommodations at their base. Commuting pilots have various options available to them for residence while at their base, including shared apartments. The cost of an occasional hotel room or sharing an apartment is not prohibitive. Our pilots have not communicated to us that they were unable to find an affordable apartment in the greater Newark, New Jersey area.

64

Commuting, like all personal time, can be managed appropriately or inappropriately. Inappropriate decisions can lead to inadequate sleep periods and fatigue. Colgan Air’s fatigue policy, scheduling policy, and pilot education and training help its pilots make the right decisions while guarding against the effects of poor decisions. Colgan Air recognizes that commuting pilots sometimes encounter difficulties getting to work in time for their rest and their assignment. Therefore, Colgan Air offers these pilots an option to call the company in advance when they know they will not be able to report on time. This policy aids the company by ensuring there is ample time to reassign a flight to a reserve pilot, and also aids the pilot by letting them know they can notify the company of a missed assignment without facing punitive action. See NTSB Public Hearing Transcript, p. 293. A pilot’s decision regarding where to live is made for personal reasons. For example, FO Shaw chose to live in Seattle to be close to family and because the commute from Seattle was easier than the commute from Norfolk. See Human Performance Group Chairman Factual Report, p. 9; Human Performance Group Chairman Factual Report Addendum, pp. 4-5. Linda Morris, mother of FO Shaw, and Michael Troy Shaw, the husband of FO Shaw, both told the NTSB that FO Shaw was not under any financial stress.25 See Human Performance Group Chairman Factual Report, p. 9; Human Performance Group Chairman Factual Report Addendum, pp. 4-5. Linda Morris said the cost of an apartment in the greater Newark area was not an issue and, as a matter of preference, FO Shaw planned to stay in hotels when needed, rather than an apartment. A captain who had flown with her also mentioned she planned to use hotels. See Human Performance Group Chairman Factual Report, p. 10. Her husband said she enjoyed flying at Colgan Air and was proud to be flying the Q400. Her long term plans were to stay in aviation and eventually upgrade to captain. Id. III.

Probable Cause

The probable cause of the accident was the flight crew’s loss of situational awareness and failure to follow Colgan Air training and procedures, which led to a loss of control of the aircraft. IV.

Contributing Causes The following contributed to the accident: 1. The flight crew’s failure to follow Colgan Air procedures and training regarding the proper response to a stick shaker.

25

We are not aware of any facts indicating FO Shaw’s choice of residence or commuting plan was influenced by her pay. On the contrary, all evidence is that pay had absolutely no bearing on these decisions.

65

2. The lack of an adequate cockpit warning system in the Q400 to warn the flight crew when a Speed Bug is set to a speed below the calculated stall warning speed. 3. The lack of an adequate warning in the Q400 Aircraft Flight Manual or Aircraft Operating Manual regarding the effect of setting a non-ice reference speed with the REF SPEEDS switch set to INCR during approach and landing. 4. The flight crew’s non-pertinent conversation during the descent and approach phase, in violation of Colgan Air’s training and procedures. V.

Recommendations

As a result of this accident, Colgan Air believes the NTSB should issue the following recommendations to the Federal Aviation Administration: 1.

Require that a Caution or Warning be required in the Q400 AOM and AFM Landing in Icing Conditions sections that states as follows: If Vref [not Vref(ice)] is used for landing, the REF SPEEDS switch must be selected to OFF or a stall warning may occur at a speed greater than Vref. or If a normal Vref speed is set in the Speed Bug when the REF SPEEDS switch is set to INCR, stall warning may occur at speeds greater than the selected Bug Speed. and If airspeed is within 20 knots of the Low Speed Cue, the airspeed must first be increased to greater than 20 knots above the Low Speed Cue before the REF SPEEDS switch is selected to INCR or a stall warning may occur.

2.

Require that the INCR REF SPEED display that appears on the Engine Display when the REF SPEEDS switch is set to INCR is either: a) changed in color to amber or red; or b) moved to a Caution Light on the Warning and Caution Lights Panel.

3.

Require that the Q400 Approach and Landing Checklist incorporate a specific positive response checklist item verifying the position of the REF SPEEDS switch and that the appropriate Vref speed is input in the Speed Bug and that landing field distance requirements are confirmed.

4.

Require that the Q400 incorporate an adequate crew warning system when a Speed Bug is set to a speed that is below the calculated stall warning speed.

66

5.

Establish a single, integrated database of pilot records that would provide airlines with real-time information about pilot qualifications and performance and would encompass not only information covered by PRIA, but also general aviation Notices of Disapproval and other Notices of Disapproval not currently covered by PRIA.

6.

In order to allow airlines to monitor and improve pilot professionalism, airlines should be permitted to use confidential groups to analyze CVR data on an anonymous, non-punitive basis.

67